Capítulo 1. Objetivo: Acessar a literatura científica para determinar o nível de evidência científica que suporta a associação da reabsorção radicular inflamatória induzida ortodonticamente (RRIIO) com diferentes sistemas de força ortodôntica. Material e Método: A busca sistemática computadorizada foi realizada nas bases de dados PubMed, Cochrane e Embase, sem restrições quanto ao ano, status ou idioma de publicação. Os critérios de seleção incluíram estudos conduzidos em no mínimo 10 pacientes submetidos ao tratamento ortodôntico com aparelhos fixos ou termoplásticos removíveis, e que apresentaram descrição do sistema de forças utilizado. Resultados: A busca eletrônica inicial das bases de dados identificou 259 artigos. Após o processo de revisão, 21 artigos preencheram os critérios de inclusão. O tamanho da amostra variou entre 10 e 73 pacientes. A maioria dos artigos foi classificada como alto nível de evidência científica e baixo risco de viés. Conclusões: A análise da literatura disponível revelou que parece existir correlação positiva entre o aumento dos níveis de força e o aumento da reabsorção radicular, bem como entre o aumento do tempo de tratamento e o aumento da reabsorção radicular. Além disso, uma pausa na movimentação ortodôntica parece ser benéfica na redução da reabsorção radicular, por facilitar o processo de reparo do cemento. A ausência de um grupo controle, de critérios de seleção dos pacientes, e de exames adequados antes e após o tratamento constituem as falhas de metodologia mais comuns no estudo da RRIIO. Capítulo 2. Objetivos: Verificar por meio da análise de elementos finitos: (1) a influência do aumento da força (25 cN e 225 cN) em três diferentes movimentos ortodônticos (intrusão, inclinação para vestibular e combinação de intrusão e inclinação) na geração de tensões críticas no ligamento periodontal (LP), considerando sua não linearidade, (2) quais tensões podem ser utilizadas como indicadores do risco à RRIIO. Material e Método: Modelo tridimensional de elementos finitos foi construído com geometria axissimétrica para simular um pré-molar inferior unirradicular e seus tecidos circundantes: cemento, LP, osso alveolar cortical e trabecular. Com exceção do LP, os demais materiais foram considerados elásticos, isotrópicos, lineares e homogêneos. Foram simulados três tipos de carregamento: intrusão pura (IP), inclinação para vestibular (IV) e combinação de forças de intrusão e inclinação (CF), e duas magnitudes de força: 25 cN e 225 cN. O deslocamento foi restrito nos nós localizados na base e na lateral do osso cortical. Quatro variáveis respostas foram analisadas: tensão de von Mises (?VM), mínima tensão principal (?3), tensão hidrostática (?H) e tensão radial (?rr). Para cada tensão foi calculado o valor nodal obtido na interface LP-cemento e a média da variável de acordo com a sua localização: lingo-cervical (LC), lingo-média (LM), lingo-apical (LA), vestíbulo-apical (VA), vestíbulo-média (VM), vestíbulo-cervical (VC). Resultados: A distribuição e os picos de ?VM (terço cervical para a intrusão e regiões LC e VA para inclinação e movimento combinado) não foram correspondentes com os locais de maior risco de RRIIO encontrados clinicamente. Maiores valores de ?3 foram obtidos: no terço apical para forças de intrusão de 25 cN e no terço cervical para 225 cN; na região VC para a inclinação e na região VA para o movimento combinado. O perfil de distribuição de ?H e ?rr foi bastante similar: os maiores valores de tensão de compressão foram encontrados no terço apical para a intrusão e nas regiões VC e LA para o movimento de inclinação e combinado. Conclusões: A tensão radial negativa foi considerada o melhor indicador do risco de RRIIO, baseado na coerência com o mecanismo biológico e nas regiões de maior risco de reabsorção para cada tipo de movimento ortodôntico. Devido ao comportamento não linear do LP, o aumento da força provocou um aumento não proporcional das tensões. Na intrusão, o aumento da tensão radial compressiva foi mais significativo para o terço apical, enquanto na inclinação e no movimento combinado para as regiões VC e LA. / Chapter 1. Objective: Assess the literature to determine which evidence level supports the association of orthodontic force system and orthodontically induced inflammatory root resorption (OIIRR). Methods: PubMed, Cochrane, and Embase databases were searched with no restrictions on year, publication status, or language. Selection criteria included human studies conducted with fixed orthodontic appliances or aligners, with at least 10 patients, and the force system well described. Results: A total of 259 articles were retrieved in the initial search. After the review process, 21 full-text articles met the inclusion criteria. Sample sizes ranged from 10 to 73 patients. Most articles were classified as having high evidence levels and low risks of bias. Conclusions: Although a meta-analysis was not performed, from the available literature, it seems that positive correlations exist between increased force levels and increased root resorption, as well as between increased treatment time and increased root resorption. Moreover, a pause in tooth movement seems to be beneficial in reducing root resorption because it allows the resorbed cementum to heal. The absence of a control group, selection criteria of patients, and adequate examinations before and after treatment are the most common methodology flaws in studies of OIIRR. Chapter 2. Objectives: Verify by the finite element analysis: (1) The influence of the loading magnitude (25 cN and 225 cN) in three different orthodontic movements (intrusion, tipping and combination of both) on the stresses generation within the periodontal ligament (PDL) considering its nonlinearity. (2) Which stress criteria can be used as indicators of root resorption risk. Material and Method: Three-dimensional finite element model was constructed with axisymmetric geometry to simulate a single-rooted lower premolar and its surrounding tissues: cementum, PDL, cortical and trabecular alveolar bone. Except for the PDL, the remaining materials were considered elastic, isotropic, linear and homogeneous. Pure intrusion (PI), buccal tipping (BT) and combination of intrusion and tipping forces (CF), applied with two force magnitudes (25 cN and 225 cN), were simulated. The model displacement was restricted at the base and the lateral of cortical bone. Four stress criteria were analyzed: von Mises (?VM), minimum principal stress (?3), hydrostatic stress (?H) and radial stress (?rr). For each criteria it was calculated the nodal stress obtained at the PDL-cementum interface and its average value according to the location: lingual-cervical (LC), lingual-middle (LM), lingual-apical (LA), buccal-apical (BA), buccal-middle (BM), and buccocervical (BC). Results: The distribution and peak of ?VM (cervical third for intrusion, LC and BA regions for buccal tipping and combined movements) did not correspond to the zones where resorption occurs clinically. The highest average values of ?3 were obtained: at the apical third under 25 cN and at the cervical third under 225 cN of intrusion forces; at the BC region under tipping forces, and at the BA region under combined forces. The distribution pattern of ?H and ?rr was very similar: the highest compressive stress values was found at the apical third under intrusion forces and at the BC and LA regions under tipping and combined forces. Conclusions: The radial stress can be considered as the best indicator to predict the OIIRR risk, based on its consistency with the biological mechanism and according to the zones where resorption occurs clinically depending on the type of orthodontic movement. Due to the PDL nonlinearity, the increase of force magnitude caused a non-linear increase of the stress values. For intrusion, the increase of radial stresses was more significant at the apical third, while for buccal tipping and combined movement it was more significant for the BC and LA regions.

Finite element models were created in Abaqus 6.14 to characterize the rotational stiffness of shallowly embedded column-foundation connections. Scripts were programmed to automate the model generation process and allow study of multiple independent variables, including embedment length, column size, baseplate geometry, concrete modulus, column orientation, cantilever height, and applied axial load. Three different connection types were investigated: a tied or one part model; a contact-based model; and a cohesive-zone based model. Cohesive-zone modeling was found to give the most accurate results. Agreement with previous experimental data was obtained to within 27%. Baseplate geometry was found to affect connection stiffness significantly, especially at lower embedment depths. The connection rotational stiffness was found to vary only slightly with cantilever height for typical column heights. Results from varying other parameters are also discussed.

The current assembly procedure to install a trunnion and hub into a bascule bridge girder involves cooling the trunnion in liquid nitrogen and shrink fitting it into the hub. The resulting trunnion-hub assembly is then allowed to warm to room temperature. Next, the trunnion-hub assembly is cooled in the liquid nitrogen and shrunk fit into the girder. The cooling of the trunnion does not cause any problems, however, when the trunnion-hub assembly is cooled in the liquid nitrogen, the hub experiences a large thermal shock. These thermal shocks induce large stresses into the hub, which has been known to cause it to crack. This study investigates an innovative assembly procedure to install the trunnion-hub assembly into a bascule bridge girder. To avoid cooling the trunnion-hub assembly, the girder was heated instead. Laboratory testing and finite element analysis were used to determine if the girder could reasonably be heated to install the trunnion-hub assembly. An experiment was conducted to analyze the heating process that will be used. A rectangular steel plate ( 60"x60x"0.75") was used to model the girder in the lab. Inductance-heating coils were used to heat the steel plate to 350°F. The heating process was recorded using a data acquisition system with thermocouples and strain gages. ANSYS was the finite element analysis (FEA) program that was used to model the heating process of the plate. The FEA results from ANSYS were compared with the experimental results. This confirmed the parameters of the finite element analysis were correct. Those parameters were then used to model a full-scale girder. The feasibility of heating the girder was determined from the finite element analysis results. It was determined that heating the girder with 2250 BTU over min for 90 minutes, was sufficient energy for the assembly procedure to work. The girder was heated to a maximum temperature of 350°F and a 0.015" clearance was created for the assembly of the trunnion-hub. The finite element analysis of the girder showed that the placement of the heating coils on the girder was critical. Therefore this innovative assembly procedure can easily be accomplished, however, each girder must first be analyzed to determine the optimal heating configuration.

Placing a geomembrane liner in the core of a dam is an alternative construction technique to traditional clay core types. This study aims to assess the performance of such internal geomembrane sealing systems in an earthen dam. Two-dimensional (2D) numerical analysis was performed to evaluate leakage through defective seams within an earthen dam. Five possible applications of internal geomembrane systems were initially modeled to locate the zero-pressure lines in an earthen dam. Then, another application where the geomembrane is placed on the upstream face was modeled to compare the upstream and internal geomembrane systems. The results of this study show that use of a geomembrane system, either upstream or internal, significantly decreases the pore pressure at the downstream face of the earthen dam. In addition, limit equilibrium analysis was performed to evaluate the effects of leakage through defects in geomembranes on the dam stability. The stability analyses for the upstream and downstream slopes were performed for three loading conditions: (1) end of construction, (2) long-term, and (3) rapid drawdown. The frequencies and locations of defective seams had a significant impact on the factors of safety of the downstream slope. It is shown that, in the case of upstream geomembrane systems, the factor of safety for the downstream slope has the highest value when the geomembrane hole occurs at a relatively lower location. On the other hand, in the case of internal geomembrane systems, the highest factor of safety occurs when the geomembrane hole is at a higher location. Additionally, rapid drawdown simulations show that the upstream slope of an embankment dam must be flat enough to overcome the upstream stability issues when geomembranes are placed within embankment dams. This study not only showed the advantages of using a geomembrane in the core of a dam as an impervious lining system but also provided comparative information on the performance of internal and upstream geomembrane systems with respect to the stability in earthen dams.

Prohibitive cost and large size of conventional angular rate sensors have limited their use to large scale aeronautical applications. However, the emergence of MEMS technology in the last two decades has enabled angular rate sensors to be fabricated that are orders of magnitude smaller in size and in cost. The reduction in size and cost has subsequently encouraged new applications to emerge, but the accuracy and resolution of MEMS angular rate sensors will have to be greatly improved before they can be successfully utilised for such high end applications as inertial navigation. MEMS angular rate sensors consist of a vibratory structure with two main resonant modes and high Q factors. By means of an external excitation, the device is driven into a constant amplitude sinusoidal vibration in the first mode, normally at resonance. When the device is subject to an angular rate input, Coriolis acceleration causes a transfer of energy between the two modes and results in a sinusoidal motion in the second mode, whose amplitude is a measure of the input angular rate. Ideally the only coupling between the two modes is the Coriolis acceleration, however fabrication imperfections always result in some cross stiffness and cross damping effects between the two modes. Much of the previous research work has focussed on improving the physical structure through advanced fabrication techniques and structural design; however attention has been directed in recent years to the use of control strategies to compensate for these structural imperfections. The performance of the MEMS angular rate sensors is also hindered by the effects of time varying parameter values as well as noise sources such as thermal-mechanical noise and sensing circuitry noise. In this thesis, MEMS angular rate sensing literature is first reviewed to show the evolu- tion of MEMS angular rate sensing from the basic principles of open-loop operation to the use of complex control strategies designed to compensate for any fabrication imperfections and time-varying effects. Building on existing knowledge, a novel adaptively controlled MEMS triaxial angular rate sensor that uses a single vibrating mass is then presented. Ability to sense all three components of the angular rate vector with a single vibrating mass has advantages such as less energy usage, smaller wafer footprint, avoidance of any mechanical interference between multiple resonating masses and removal of the need for precise alignment of three separate devices. The adaptive controller makes real-time estimates of the triaxial angular rates as well as the device cross stiffness and cross damping terms. These estimates are then used to com- pensate for their effects on the vibrating mass, resulting in the mass being controlled to follow a predefined reference model trajectory. The estimates are updated using the error between the reference model trajectory and the mass's real trajectory. The reference model trajectory is designed to provide excitation to the system that is sufficiently rich to enable all parameter estimates to converge to their true values. Avenues for controller simplification and optimisation are investigated through system simulations. The triaxial controller is analysed for stability, averaged convergence rate and resolution. The convergence rate analysis is further utilised to determine the ideal adaptation gains for the system that minimises the unwanted oscillatory behaviour of the parameter estimates. A physical structure for the triaxial device along with the sensing and actuation means is synthesised. The device is realisable using MEMS fabrication techniques due to its planar nature and the use of conventional MEMS sensing and actuation elements. Independent actuation and sensing is achieved using a novel checkerboard electrode arrangement. The physical structure is refined using a design automation process which utilises finite element analysis (FEA) and design optimisation tools that adjust the design variables until suitable design requirements are met. Finally, processing steps are outlined for the fabrication of the device using a modified, commercially available polysilicon MEMS process.

The handling of variability effects in structural models is a natural and necessary extension of deterministic analysis techniques. In the context of finite element and uncertainty modelling, the stochastic finite element method (SFEM), grouping the perturbation SFEM, the spectral SFEM and the Monte-Carlo simulation, has by far received the major attention. <br> The present work focuses on second moment approaches, in which the first two statistical moments of the structural response are estimated. Due to its efficiency for handling problems involving low variability levels, the perturbation method is selected for characterising the propagation of the parameter variability from an uncertain dynamic model to its structural response. A dynamic model excited by a time-harmonic loading is postulated and the extension of the perturbation SFEM to the frequency domain is provided. This method complements the deterministic analysis by a sensitivity analysis of the system response with respect to a finite set of random parameters and a response surface in terms of a Taylor series expansion truncated to the first or second order is built. Taking into account the second moment statistical data of the random design properties, the response sensitivities are appropriately condensed in order to obtain an estimation of the response mean value and covariance structure. <br> In order to handle a wide definition of variability, a computational tool is made available that is able to deal with material variability sources (material random variables and fields) as well as shape uncertainty sources. This second case requires an appropriate shape parameterisation and a shape design sensitivity analysis. The computational requirements of the tool are studied and optimised, by reducing the size of the random dimension of the problem and by improving the performances of the underlying deterministic analyses. In this context, modal approaches, which are known to provide efficient alternatives to direct approaches in frequency domain analyses, are developed. An efficient hybrid procedure, coupling the perturbation and the Monte-Carlo simulation SFEM, is proposed and analysed. <br> Finally, the developed methods are validated, by resorting mainly to the Monte-Carlo simulation technique, on different numerical applications: a cantilever beam structure, a plate bending problem (involving a 3-dimensional model), an articulated truss structure and a problem involving a plate with a random flatness default. The propagation of the model uncertainty in the response FRFs and the effects involved by random field modelling are examined. Some remarks are stated pertaining to the influence of the parameter PDF in simulation-based methods. <br> <br> La gestion de la variabilité présente dans les modèles structuraux est une extension naturelle et nécessaire des techniques de calcul déterministes. En incorporant la modélisation de l'incertitude dans le calcul aux éléments finis, la méthode des éléments finis stochastiques (groupant l'approche perturbative, l'approche spectrale et la technique de simulation Monte-Carlo) a reçu une large attention de la littérature scientifique. <br> Ce travail est orienté sur les approches dites de second moment, dans lesquelles les deux premiers moments statistiques de la réponse de la structure sont estimés. De par son aptitude à traiter des problèmes caractérisés par de faibles niveaux de variabilité, la méthode perturbative est choisie pour propager la variabilité des paramètres d'un modèle dynamique incertain sur sa réponse. Un modèle sous chargement dynamique harmonique est supposé et l'extension dans le domaine fréquentiel de l'approche perturbative est établie. Cette méthode complète l'analyse déterministe par une analyse de sensibilité de la réponse du système par rapport à un ensemble fini de variables aléatoires. Une surface de réponse en termes d'un développement de Taylor tronqué au premier ou second ordre peut alors être écrit. Les sensibilités de la réponse sont enfin condensées, en tenant compte des propriétés statistiques des paramètres de design aléatoires, pour obtenir une estimation de la valeur moyenne et de la structure de covariance de la réponse. <br> Un outil de calcul est développé avec la capacité de gestion d'une définition large de la variabilité: sources de variabilité matérielle (variables et champs aléatoires) ainsi que géométrique. Cette dernière source requiert une paramétrisation adéquate de la géométrie ainsi qu'une analyse de sensibilité à des paramètres de forme. Les exigences calcul de cet outil sont étudiées et optimisées, en réduisant la dimension aléatoire du problème et en améliorant les performances des analyses déterministes sous-jacentes. Dans ce contexte, des approches modales, fournissant une alternative efficace aux approches directes dans le domaine fréquentiel, sont dérivées. Une procédure hybride couplant la méthode perturbative et la technique de simulation Monte-Carlo est proposée et analysée. <br> Finalement, les méthodes étudiées sont validées, principalement sur base de résultats de simulations Monte-Carlo. Ces résultats sont relatifs à plusieurs applications numériques: une structure poutre-console, un problème de flexion de plaque (modèle tridimensionnel), une structure en treillis articulé et un problème de plaque présentant un défaut de planéité aléatoire. La propagation de l'incertitude du modèle dans les fonctions de réponse fréquentielle ainsi que les effets propres à la modélisation par champs aléatoires sont examinés. Quelques remarques relatives à l'influence de la loi de distribution des paramètres dans les méthodes de simulation sont évoquées.

The stiffness and frequency properties of tensegrity structures are functions of the pre-stress, topology, configuration, and axial stiffness of the elements. The tensegrity structures considered are tensegrity booms, tensegrity grids, and tensegrity power lines. A study has been carried out on the pre-stress design. It includes (i) finding the most flexible directions for different pre-stress levels, (ii) finding the pre-stress pattern which maximizes the first natural frequency. To find the optimum cross-section areas of the elements for triangular prism and Snelson tensegrity booms, an optimization approach is utilized. A constant mass criterion is considered and the genetic algorithm (GA) is used as the optimization method. The stiffness of the triangular prism and Snelson tensegrity booms are modified by introducing actuators. An optimization approach by means of a GA is employed to find the placement of the actuators and their minimum length variations. The results show that the bending stiffness improves significantly, but still an active tensegrity boom is less stiff than a passive truss boom. The GA shows high accuracy in searching the non-structural space. The tensegrity concept is employed to design a novel transmission power line .A tensegrity prism module is selected as the building block. A complete parametric study is performed to investigate the influence of several parameters such as number of modules and their dimensions on the stiffness and frequency of the structure. A general approach is suggested to design the structure considering wind and ice loads. The designed structure has more than 50 times reduction of the electromagnetic field and acceptable deflections under several loading combinations. A study on the first natural frequencies of Snelson, prisms, Micheletti, Marcus and X-frame based tensegrity booms has been carried out. The result shows that the differences in the first natural frequencies of the truss and tensegrity booms are significant and not due to the number of mechanisms or pre-stress levels. The tensegritybooms of the type Snelson with 2 bars and prism with 3 bars have higher frequencies among tensegrity booms. / <p>QC 20120904</p>

In mining industry, waste storage is a very prominent issue / in this respect, safety of storage structures is one of the leading problems in the industry. Most of the tailings dams require remedial measures, throughout their lifespan to increase their reliability. The objective of the study is to investigate stability problems of formerly constructed but newly raised Eti Copper Mine tailings dam and alternative dam geometries for future raises. Plenty of methods were developed to analyze the reliability of structures / limit equilibrium methods, finite element methods and finite difference methods are among them. In this case, stability of the dam was analyzed with finite element method under static loading conditions. In order to determine input parameters properly, disturbed samples obtained at the field investigations were used. For this purpose, several laboratory experiments were conducted to determine natural moisture content, grain size distribution, specific gravity, Atterberg limits, maximum dry density and shear strength parameters of tailings and embankment material.

The high specific bending stiffness of sandwich structures can with advantage be used in vehicles to reduce their weight and thereby potentially also their fuel consumption. However, the structure must not only meet the in-service requirements but also provide sufficient protection of the vehicle passengers in a crash situation. The in-plane compressive response of sandwich panels is investigated in this thesis, with the objective to develop a methodology capable of determining if the structural response is likely to be favourable in an energy absorption perspective. Experiments were conducted to identify possible initial failure and collapse modes. The initial failure modes of sandwich panels compressed quasi-statically in the in-plane direction were identified as global buckling, local buckling (wrinkling) and face sheet fracture. Global buckling promotes continued folding of the structure when compressed beyond failure initiation. Face sheet fracture and wrinkling can promote collapse in the form of unstable debond crack growth, stable end-crushing or ductile in-plane shear collapse. Both the unstable debond crack propagation and the stable end-crushing are related to debond crack propagation, whereas the ductile in-plane shear mode is related to microbuckling of the face sheets. The collapse behaviour of sandwich configurations initially failing due to wrinkling or face sheet fracture was investigated, using a finite element model. The model was used to determine if the panels were likely to collapse in unstable debond propagation or in a more stable end-crushing mode, promoting high energy absorption. The collapse behaviour is mainly governed by the relation between the fracture toughness of the core and the bending stiffness and strength of the face sheets. The model was successfully used to design sandwich panels for different collapse behaviour. The proposed method could therefore be used in the design process of sandwich panels subjected to in-plane compressive loads.During a crash situation the accelerations on passengers must be kept below life threatening levels. The extreme peak loads in the structure must therefore be limited. This can be achieved by different kind of triggering features.Panels with either chamfered face sheets or with grooves on the loaded edges were investigated in this thesis. The peak load was reduced with panels incorporating either of the two triggering features. Another positive effect was that the plateau load following failure initiation was increased by the triggers. This clearly illustrates that triggers can be used to promote favourable response in sandwich panels. Vehicles are harmful to the environment not only during in-serve use, but during their entire life-cycle. By use of renewable materials the impact on the environment can be reduced. The in-plane compressive response of bio-based sandwich panels was therefore investigated. Panels with hemp fibre laminates showed potential for high energy absorption and panels with a balsa wood core behaved particular well. The ductile in-plane shear collapse mode of these panels resulted in the highest energy absorption of all investigated sandwich configurations. / QC 20100728

Mechanical pulping for producing pulps from softwood suitable for printing grade papers, like news, is a highly energy-intensive process consuming around 2000 kWh/t in electrical energy. Due to increasing energy costs and environmental issues there is a high demand for decreasing this energy consumption. The mechanical treatment of wet wood pieces in a refiner, in the mechanical pulp plant, is a complex mechanical loading. This is a process occurring between rotating discs at high speed and temperatures of 140 °C - 160 °C, where by means of shear and compression forces the fibres are separated and then made flexible, fibrillated and collapsed for good bonding ability. In this process also fines are created giving the optical properties of the paper. In mechanical pulping only a fraction of the applied energy is used for the structural changes of the wood material. Thus fundamental studies of the loading modes of wood under refining conditions and in particular under combined shear and compression loading are desired to gain more information regarding the possibility of affecting the mechanical pulping in an energy efficient way. The possibilities to study the behaviour of wood under a combined shear and compression load were in this thesis investigated using two methods: the Iosipescu shear test and the Arcan shear test. In both apparatus different combinations of shear and compression load were achieved by different rotations of the shear test device itself. Measurements with the Iosipescu device on a medium density fibreboard showed good agreement between experimental results and numerical simulations. Finite element analysis on wood showed, however, that with the use of a homogeneous material in the model the level of strain reached would be ten times smaller than experimentally measured. This fact is probably due to the honeycomb structure of the wood cells that allows for different local deformations that could not be represented by a continuous material model. Thus to study the deformations on the fibre level of wood an experimental equipment that uses smaller samples was needed. With a modified Arcan shear device such deformations under combined shear and compression load and in pure compression were possible showing different deformation patterns. During pure compression the cell walls bend in a characteristic “S” shape, independently of the shape of the fibre cells and their cell wall thickness. Under combined shear and compression, however, mainly the corners of the fibre cells deform giving a “brick” shape to the cells. In a second deformation performed in compression, the fibre cells follow the same deformation pattern as given by the first deformation type whether in compression or in combined shear and compression. The interpretation is that permanent defects in the cells themselves are introduced already in the first load cycle of the wood samples. The energy used under the different loading conditions showed that the first deformation required the largest amount of energy, for all loading conditions. The deformation in compression required larger amounts of energy than the deformation in combined loads. For subsequent deformations less energy was needed for compression if a combined load had preceded it. Due to the fact that less energy is needed to start to deform wood in combined load than under compression load, the application of a combined load as a first cycle may thus be a way to permanently deform fibres using less energy. To investigate the critical parameters determining the permanent deformation of cells, a finite element model of a network of twelve cells was developed. Special care was given to the material properties to study how the variation of the fibril angle in the different layers affects the deformation pattern of the wood fibres under the different loading conditions. The model shows that whether modelled as homogeneous linear isotropic material or as an orthotropic material defined for every layer of the cells wall, no difference in the deformation of the network of the fibres was achieved. It is probable that the deformation type is more determined by the geometry of the fibres themselves than by their material properties / QC 20101005