Abstract Fretting is present where the contacting surfaces of mechanical parts are subjected to rubbing and an alternating stress, resulting in fatigue in the material. This is the situation between the hub and shaft in a shrink-fitted assembly. In practice, fretting is found in the same assembly with normal fatigue, and it occurs in the axle in a shrink-fit between an axle and a hub, while normal fatigue cracks are found outside the assembly. Fretting phenomena have been investigated by many authors in laboratories, but the dimensioning criteria for shrink-fit assemblies are insufficient for the construction engineer. Fretting causes a considerable reduction in the fatigue strength of a shrink-fit assembly, and failures caused by fretting are as numerous as failures resulting from normal fatigue. The purpose of this investigation was to examine the effect of hub material, contact pressure, slip amplitude and shaft geometry on fatigue strength. The goal of this investigation was to determine an optimal contact pressure and a favourable fillet radius and axle diameter ratio at which fretting failure can be avoided and maximum normal fatigue strength will be obtained. The torsional fatigue strength of shrink-fitted shaft couplings was estimated using tests that varied the material of the hub, the contact pressure, the geometry of the shaft and the torsional stress amplitude of the shrink-fitted assembly. Based on the information obtained from the test, aluminium as a hub material appears to produce little damage to a steel shaft, whereas bronze appears to cause damage and fatigue. Cast iron and steel caused a medium amount of damage. To increase contact pressure at the shoulder, hub overhang past the shoulder was used. These experiments showed that increasing the contact pressure decreased the slip amplitude, thereby reducing fretting. With low contact pressure, shaft fracturing began from fretting fatigue inside the hub, but with high contact pressure the shaft fractured at the fillet due to normal fatigue. Selecting the fillet radius according to the contact pressure makes it possible to dimension the shrink-fit shaft assembly to prevent fretting. The shrink-fitted shaft should be designed according to the normal fatigue limit, because fretting fatigue may occur when the number of load cycles exceeds 2×107. Fretting can be prevented by using a sufficiently high contact pressure and by choosing optimal ratios for the fillet radius and the diameter. To avoid fretting, the slip amplitude should be under 3 μm. This is achieved by using a contact pressure of over 100 N/mm2, calculated according to Lamé's theory. A method for dimensioning a shrink-fitted shaft with respect to fretting fatigue was presented based on a specific geometry (Ø 50 mm) and the materials used in the tests. The method was applied in verifying the fretting fatigue of a shrink-fitted shaft with a diameter of 300 mm. / Tiivistelmä Kitkaväsyminen esiintyy yleisesti, kun koneenosien kontaktipintoihin kohdistuu edestakainen jännitysamplitudi hankaavan liikkeen lisäksi johtaen materiaalin väsymiseen. Tällainen tilanne vallitsee navan ja akselin välissä kutistusliitoksessa. Kitkaväsyminen esiintyy käytännössä samassa kutistusliitoksessa kuin normaali väsyminenkin ja se ilmenee navan ja akselin välissä yleensä akselissa, kun taas tavanomainen väsyminen tapahtuu akselissa liitoksen ulkopuolella. Kitkaväsymistä on tutkittu paljonkin, mutta kutistusliitoksen mitoituskriteerit kitkaväsymisen osalta ovat puutteelliset. Kitkaväsyminen aiheuttaa merkittävän väsymislujuuden heikkenemisen sekä täten väsymisiän alenemisen kutistusliitoksessa ja sen aiheuttamat vauriot ovat määrältään samaa luokkaa tavanomaisen väsymisen kanssa. Tämän tutkimuksen tarkoituksena oli selvittää napamateriaalin, liitospaineen, liukuma-amplitudin ja geometrian vaikutus väsymislujuuteen. Tavoitteena oli määrittää optimaalinen liitospaine sekä sopivat olakkeen pyöristyssäde ja akselisuhde, jotta kitkaväsyminen voidaan välttää ja saavuttaa maksimaalinen normaali väsymislujuus. Kitkaväsymislujuutta väännön suhteen testattiin varioimalla napamateriaalia, liitospainetta, geometriaa ja jännitysamplitudia. Testien perusteella alumiininen napamateriaali sopi hyvin teräsakselin kanssa, kun taas pronssi aiheutti akseliin pintavaurion ja sitä kautta väsymisilmiön. Teräs- ja valurautanapa olivat näiden kahden materiaalin välissä. Liitospaineen nostamiseksi olakkeen reunalla käytettiin navan ylitystä olakkeen yli. Kokeet osoittivat kitkaväsymisen vähenevän korkeammilla liitospaineilla liukuman aletessa. Alhaisella liitospaineella akselin vaurioituminen alkoi kitkaväsymisenä navan sisältä, mutta hyvin korkealla liitospaineella murtuminen tapahtui olakkeesta tavanomaisena väsymisenä. Valitsemalla pyöristyssäde liitospaineen perusteella kutistusliitos on mitoitettavissa kitkaväsymistä vastaan. Kutistusliitos tulisi mitoittaa tavanomaisen väsymisen mukaan, koska kitkaväsymismurtuma voi tapahtua kuormanvaihtoluvulla yli 2×07. Kitkaväsyminen on ehkäistävissä käyttämällä riittävän korkeata liitospainetta sekä sopivaa olakkeen pyöristyssädettä yhdistyneenä oikeaan akselisuhteeseen. Kitkaväsyminen estyy kun liukuma-amplitudi on alle 3 μm. Tämä on saavutettavissa liitospaineella yli 100 N/mm2 laskettuna Lamén teorian mukaan. Tutkimuksessa on esitetty mitoitusmenetelmä kitkaväsymisen suhteen perustuen Ø 50 mm akselilla sekä käytetyillä materiaaleilla tehtyihin testeihin. Menetelmää on sovellettu kitkaväsymisen tarkasteluun kutistusliitokseen, jonka akselin halkaisija on 300 mm.

Structural steel design standards recognize lateral torsional buckling as a failure mode governing the capacity of long span unsupported beams with wide flange cross-sections. Standard solutions start with the closed form solution of the Vlasov thin-walled beam theory for the case of a simply supported beam under uniform moments, and modify the solution to accommodate various moment distributions through moment gradient expressions. The Vlasov theory solution is based on the assumption that cross-sectional distortional effects have a negligible effect on the predicted elastic critical moment. The present study systematically examines the validity of the Vlasov assumption related to cross-section distortion through a parametric study. A series of elastic shell finite element eigen-value buckling analyses is conducted on simply supported beams subject to uniform moments, linear moments and mid span point loads as well as cantilevers subject to top flange loading acting at the tip. Cross-sectional dimensions are selected to represent structural steel cross-section geometries used in practice. Particular attention is paid to model end connection details commonly used in practice involving moment connections with two pairs of stiffeners, simply supported ends with a pair of transverse stiffeners, simply supported ends with cleat angle details, and built in fixation at cantilever roots. The critical moments obtained from the FEA are compared to those based on conventional critical moment equations in various Standards and published solutions. The effects of web slenderness, flange slenderness, web height to flange width ratio, and span to height ratios on the critical moment ratio are systematically quantified. For some combinations of section geometries and connection details, it is shown that present solutions derived from the Vlasov theory can overestimate the lateral torsional buckling resistance for beams.

A theoretical study is conducted for the lateral torsional buckling of wooden beam-deck assemblies consisting of twin beams braced by tongue-and-groove decking at the top. Two models are developed, each with a series of analytical and numerical solutions formulated. The first model targets twin-beam-deck assemblies where deck boards and other components are detailed to provide full continuous lateral restraint while the second model is built for situations where the beams are allowed to sway laterally and the relative lateral movement between the beams is partially restrained by the deck boards. In the first model, focus is on wind uplift while in the second model, both gravity and uplift loading scenarios are investigated. In the first model, an energy method is adopted and the principle of stationary potential energy is evoked to formulate closed-form solutions, energy-based solutions and a finite element solution. The validity of the present solutions is verified against a finite element based ABAQUS model. Similarly, a family of solutions is developed under the sway model and verified against the ABAQUS. Parametric studies are conducted for both models to examine the effects of various variables on the buckling capacity. A comparative investigation on the behavioral difference between the two models under ABAQUS is also presented. Overall, the restraining effects of deck boards bracing either on the beam compression or tension side is observed to have a significant influence on the lateral torsional buckling capacity of the twin-beam-deck assemblies.

An analytical and numerical investigation is conducted for the lateral torsional buckling analysis of wooden beam with a mid-span lateral brace subjected to symmetrically distributed loading. Two models are developed; one for the case of a rigid brace and another one for the case of a flexible brace. The analytical solutions are based on the principle of stationary potential energy and a Fourier expansion of the buckling displacement fields and bending moments. The validity of both models are verified against 3D finite element analyses in ABAQUS. Where applicable, verifications were also conducted against available solutions from previous studies. Parametric studies were conducted to investigate the effect of geometric and material parameters on the critical moments. The results indicate the presence of two separate groups of potential buckling modes, symmetric and anti-symmetric, with fundamentally different behavioural characteristics. The governing buckling mode is shown to depend on the bracing height, load height and lateral brace stiffness. The study shows that beyond a certain threshold bracing height, the critical moment is governed by the antisymmetric mode of buckling. Also, above a certain optimum bracing stiffness, no increase is observed in the critical moments. The models developed are used to construct a comprehensive database of parametric investigations which are then employed for developing simplified equations for determining the threshold heights, associated critical moments, and optimum stiffness.

Engineered wood I-joists have grown in popularity as flooring and roofing structural systems in the past 30 years, replacing solid sawn lumber joists. Typical wood I-joists are manufactured with a very slender section, which is desirable to achieve higher flexural capacities and longer spans; however, this makes them susceptible to lateral torsional buckling failure. Continuous beam spans and uplift forces on roof uplift are potential scenarios where lateral instability can occur and reflects the need to investigate the lateral torsional buckling behavior of wood I-joists. Within this context, the present study conducts an experimental investigation on the material properties and the critical buckling load of 42 wood I-joist specimens. A 3D finite element model is built using the experimentally determined material parameters to effectively predict the observed buckling behavior of the specimens while also accounting for initial imperfections in the joists. The adequacy of other analytical models to predict the critical buckling load of wood I-joists are also investigated. It is demonstrated that the American design standard underestimates the critical buckling load of wood I-joists while the classical theory provides an adequate estimate of the buckling capacity. Furthermore, the effects of initial imperfections on the lateral torsional buckling behavior are discussed. The developed and verified FE model is used to reproduce the nonlinear buckling behavior of the wood I-joist and also to provide an accurate estimate of the lateral torsional buckling capacity using the linear buckling analysis.

Existing design provisions in design standards and conventional analysis methods for structural steel members are based on the simplifying kinematic Vlasov assumption that neglects cross-sectional distortional effects. While the non-distortional assumption can lead to reasonable predictions of beam static response and buckling strength in common situations, past work has shown the inadequacy of such assumption in a number of situations where it may lead to over-predicting the strength of the members. The present study thus develops a series of generalized theories/solutions for the static analysis and buckling analysis of steel members with wide flange cross-sections that capture distortional effects of the web. Rather than adopting the classical Vlasov assumption that postulates the cross-section to move and rotate in its own plane as a rigid disk, the present theories assume the web to be flexible in the plane of the cross-section and thus able to bend laterally, while both flanges to move as rigid plates within the plane of the cross-section to be treated as Euler-Bernouilli beams. The theories capture shear deformation effects in the web, as well as local and global warping effects. Based on the principle of minimum potential energy, a distortional theory is developed for the static analysis of wide flange steel beams with mono-symmetric cross-sections. The theory leads to two systems of differential equations of equilibrium. The first system consists of three coupled equilibrium differential equations that characterize the longitudinal-transverse response of the beam and the second system involves four coupled equilibrium differential equations of equilibrium and characterizes the lateral-torsional response of the beam. Closed form solutions are developed for both systems for general loading. Based on the kinematics of the new theory, two distortional finite elements are then developed. In the first element, linear and cubic Hermitian polynomials are employed to interpolate displacement fields while in the second element, the closed-form solutions developed are adopted to formulate special shape functions. For longitudinal-transverse response the elements consist of two nodes with four degree of freedom per node for longitudinal-transverse response and for lateral-torsional response, the elements consist of two nodes with eight degrees of freedom per node. The solution is able to predict the distortional deformation and stresses in a manner similar to shell solutions while keeping the modeling and computational effort to a minimum. Applications of the new beam theory include (1) providing new insights on the response of steel beams under torsion whereby the top and bottom flanges may exhibit different angles of twist, (2) capturing the response of steel beams with a single restrained flange as may be the case when a concrete slab provides lateral and/or torsional restraint to the top flange of a steel beam, and (3) modelling the beneficial effect of transverse stiffeners in reducing distortional effects in the web. The second part of the study develops a unified lateral torsional buckling finite element formulation for the analysis of beams with wide flange doubly symmetric cross-sections. The solution captures several non-conventional features. These include the softening effect due to web distortion, the stiffening effect induced by pre-buckling deformations, the pre-buckling nonlinear interaction between strong axis moments and axial forces, the contribution of pre-buckling shear deformation effects within the plane of the web, the destabilizing effects due to transverse loads being offset from the shear centre, and the presence of transverse stiffeners on web distortion. Within the framework of the present theory, it is possible to evoke or suppress any combination of the features and thus isolate the individual contribution of each effect or quantify the combined contributions of multiple effects on the member lateral torsional capacity. The new solution is then applied to investigate the influence of the ratios of beam span-to-depth, flange width-to-thickness, web height-to-thickness, and flange width-to-web height on the lateral torsional buckling strength of simply supported beams and cantilevers. Comparisons with conventional lateral torsional buckling solutions that omit distortional and pre-buckling effects quantify the influence of distortional and/or pre-buckling deformation effects. The theory is also used to investigate the influence of P-delta effects of beam-columns subjected to transverse and axial forces on their lateral torsional buckling resistance. The theory is used to investigate the load height effect relative to the shear centre. Comparisons are made with load height effects as predicted by non-distortional buckling theories. The solution is adopted to quantify the beneficial effect of transverse stiffeners in controlling/suppressing web distortion in beams and increasing their buckling resistance.

Modern 'more-electric' aircraft demand increased levels of electrical power as non-propulsive power systems are replaced with electrical equivalents. This electrical power is provided by electrical generators, driven via a mechanical transmission system, from a rotating spool in the gas turbine core. A wide range of electrical loads exist throughout the aircraft, which may be pulsating and high powered, and this electrical power demand is transferred though the generators to produce a torque load on the drivetrain. The mechanical components of the drivetrain are designed for minimum mass and so are susceptible to fatigue, therefore the electrical loading existing on modern airframes may induce fatigue in key mechanical components and excite system resonances in both mechanical and electrical domains. This electro-mechanical interaction could lead to a reduced lifespan for mechanical components and electrical network instability.This project investigates electro-mechanical interaction in the electrical power offtake from large diameter aero gas turbines. High fidelity modelling of the drivetrain, and generator, allow the prediction of system resonances for a generic gas turbine-generator system. A Doubly-Fed Induction Generator (DFIG) is considered and modelled. DFIGs offer opportunities due to their fast dynamics and their ability to decouple electrical and mechanical frequencies (e.g. enabling a constant frequency electrical system with a variable speed mechanical drive). A test platform is produced which is representative of a large diameter gas turbine and reproduces the electro-mechanical system behaviour. The test platform is scaled with respect to speed and power but maintains realistic sizing between component dimensions which include: a gas turbine mechanical spool emulation, transmission driveshafts and gearbox, and accessory loads such as a generator. This test platform is used to validate theoretical understanding and suggest alternative mechanical configurations, and generator control schemes, for the mitigation of electro-mechanical interaction.The novel use of a DFIG and an understanding of electro-mechanical interaction allow future aircraft designs to benefit from the increased electrification of systems by ensuring that sufficient electrical power can be provided by a robust gas turbine-generator system.

The ability to withstand a short circuit is the most essential feature of a power transformer. The most important reason to design short-circuits proof transformers is to ensure the reliability of the power grid (avoiding black outs etc.) and safety (fire and explosion in case of failure). During short circuit, the most effected winding is the LV winding due to the flow high currents even during the normal working condition. So during a short circuit large forces are generated which act on the winding and these forces can reach hundreds of tons in fraction of a second, so the transformer must be properly designed in order to withstand these forces or the transformer can fail in different ways. One of the possible failure modes called “Spiraling” is discussed and analyzed in this thesis. Spiraling Occurs when the LV winding twists tangentially in the opposite direction at the ends due to radial short circuit forces. From literature study the transient forces acting on the winding during a 3-phase short circuit was determined and these transient forces were used to perform simulations on the model. The axial and radial forces applied on the model were such that it has a uniform magnitude per each turn. Various analysis was performed on the model which includes the Static, Modal and Transient Structural analysis in Ansys Workbench and each analysis involved parametric analysis where the deformations and the torsional mode shapes were determined

The aim of this study was to evaluate the cyclic and torsional fatigue resistance of Nickel-Titanium rotary instruments manufactured by different thermal treatments. Were tested a total of 140 rotary instruments (n=20): Genius (GN size 25, .04 taper), Trushape (TS size 25, .06 taper), Logic (LOG size 25, .06 taper), Vortex Blue (VB size 25, .06 taper), ProTaper Gold (PTG size 25, .08 taper), Hyflex CM (HCM size 25, .06 taper) and Hyflex EDM (EDM size 25, .08 taper). Cyclic fatigue resistance testing was performed using an artificial stainless steel canal with a curvature (60° angle and 5- mm radius) located at 5 mm from the tip. The files (n=10) rotated until fracture and time was recorded in seconds. The torsional test evaluated the angular deflection and torque at failure of the instruments (n=10) at 3 mm from the tip according to ISO 3630- 1. The fractured surface of five instruments of each brand was observed by using scanning electron microscopy (SEM). Data were analysed using one-way ANOVA and Tukey tests, and the level of significance was set at 5%. The cyclic fatigue resistance value of EDM size 25, .08 taper was significantly higher than those of all instruments tested (P<0.05). The LOG size 25, .06 taper showed a higher cyclic fatigue resistance than those of GN size 25, .04 taper; TS size 25, .06 taper (P<0.05). There was no difference among the others groups. The torsional test showed that PTG size 25, .08 taper had the highest torsional strength value of all instruments tested followed by VB size 25, .06 taper and EDM size 25, .06 taper (P<0.05). The LOG size 25, .06 taper showed significant difference only with GN size 25, .04 taper (P<0.05). No difference was found among the others groups (P>0.05). In relation to angular deflection, the GN size 25, .04 taper; TS size 25, .06 taper; HCM size 25, .06 taper, and EDM size 25, .08 taper showed significantly higher values until fracture than the others groups (P<0.05). No difference was found among PTG size 25, .08 taper, LOG size 25, .06 taper, and VB size 25, .06 taper (P<0.05). The EDM size 25, .08 taper presented the highest cyclic fatigue resistance among all the tested instruments. For the torsional test, the PTG size 25, .08 taper showed highest torsional strength and lowest angular deflection values. / O objetivo deste estudo foi avaliar a resistência às fadigas cíclica e torsional de instrumentos rotatórios de Níquel - Titânio fabricados por diferentes tratamentos térmicos. Foram testados o total de 140 instrumentos (n=20): Genius (GN diâmetro 25, conicidade .04), Trushape (TS diâmetro 25, conicidade .06), Logic (LOG diâmetro 25, conicidade .06), Vortex Blue (VB diâmetro 25, conicidade .06), ProTaper Gold (PTG diâmetro 25, conicidade .08), Hyflex CM (HCM diâmetro 25, conicidade .06) e Hyflex EDM (EDM diâmetro 25, conicidade .08). O teste de resistência à fadiga cíclica foi realizado utilizando um canal artificial de aço inoxidável com curvatura (ângulo de 60° e raio de 5mm) localizada a 5 mm da ponta. Os instrumentos (n=10) foram rotacionados até a fratura e tempo foi registrado em segundos. O teste torsional avaliou a deflexão angular e torque até a falha dos instrumentos (n=10) a 3 mm da ponta de acordo com a ISO 3630-1. A superfície da fratura de 5 instrumentos de cada fabricante foi observado utilizando o microscópio eletrônico de varredura (MEV). A análise-estatística foi realizada utilizando o teste de análise de variância com um fator ANOVA e teste de Tukey, o nível de significância foi de 5%. O valor de resistência a fadiga cíclica do EDM diâmetro 25, conicidade .08 foi significantemente maior que todos os instrumentos testados (P<0.05). A LOG diâmetro 25, conicidade .06 mostrou maior resistência à fadiga cíclica que o GN diâmetro 25, conicidade .04; TS diâmetro 25, conicidade .06 (P<0.05). Não houve diferença significante entre os outros grupos. O teste torsional mostrou que PTG diâmetro 25, conicidade .08 obteve o maior valor de torque até a fratura de todos os instrumentos testados seguido por VB diâmetro 25, conicidade .06 e EDM diâmetro 25, conicidade .06 (P<0.05). O LOG diâmetro 25, conicidade .06 mostrou diferença significativa apenas com com GN diâmetro 25, conicidade .04 (P<0.05). Não houve diferença significativa entre os outros grupos (P>0.05). Em relação a deflexão angula, o GN diâmetro 25, conicidade .04; TS diâmetro 25, conicidade .06; HCM diâmetro 25, conicidade .06 e EDM diâmetro 25, conicidade .08 apresentou significantimente o maior valor até a fratura que os outros grupos (P<0.05). Nao foi encontrado diferença significativa entre PTG diâmetro 25, conicidade .08, LOG diâmetro 25, conicidade .06, e VB diâmetro 25, conicidade .06 (P<0.05). O EDM diâmetro 25, conicidade .08 apresentou a maior resistência a fadiga cíclica entre todos os instrumentos testados. Para o teste torsional, o PTG diâmetro 25, conicidade .08 apresentou o maior valor de torque e menor deflexão angular.

Continuum manipulators which are robot limbs inspired by trunks, snakes, and tentacles, represent a promising field in robotic manipulation research. They are well known for their compliance, as they can conform to the shape of objects they interact with. Furthermore, they also benefit from improved dexterity and reduced weight compared to traditional rigid manipulators. The current state of the art continuum robots typically consists of a bulky pneumatic or tendon-driven actuation system at the base, hindering their scalability. Additionally, they tend to sag due to their own weight and are weak in the torsional direction, limiting their performance under external load. This work presents an origami-inspired cable-driven continuum manipulator module that offers low-cost, light-weight, and is inherently safe for human-robot interaction. This dissertation includes contributions in the design of the modular and torsionally strong continuum robot, the motion planning and control of the system, and finally the embedded sensing to close the loop providing robust feedback.