No description available.

Diss. (sammanfattning) Stockholm : Karol. inst., 2006. / Härtill 4 uppsatser.

Das Ermüdungsverhalten von Beton wurde in der Literatur bislang sehr ausführlich anhand von zentrisch beanspruchten Zylinderprobekörpern untersucht. Untersuchungen, die sich mit zyklisch biegebeanspruchten Betonbauteilen im Druckschwellbereich beschäftigen, existieren nur in geringer Zahl. Solche Bauteile sind beispielsweise in Turmstrukturen von Windenergieanlagen vorhanden, die aktuell in großem Umfang errichtet werden. Für hochfesten Beton, wie er in diesen Bauwerken meist verwendet wird, liegen noch weniger Untersuchungsergebnisse vor. Aufgrund der hohen auftretenden Lastwechselzahlen und der großen Querschnittsabmessungen ist die versuchstechnische Untersuchung des Ermüdungsverhaltens solcher Strukturen mit großen Herausforderungen verbunden. Mit Hilfe von numerischen Simulationen kann hingegen eine Vielzahl an Parametervariationen innerhalb kürzester Zeit untersucht werden, wodurch eine beschleunigte Untersuchung des Ermüdungsverhaltens von Bauteilen möglich wird. Die dafür verwendeten Materialmodelle werden schnell sehr komplex und benötigen eine aufwendige Kalibrierung der Eingangsparameter an Bauteilversuchen. Ein Modell zur Abbildung der makroskopischen Schädigungsprozesse im Beton unter Ermüdungsbeanspruchung in Querschnitten unter Druckschwellbeanspruchung, das anhand von üblichen Zylinderversuchen kalibriert werden kann, existiert bisher nicht. In der vorliegenden Arbeit wurden diese Punkte aufgegriffen und ein Modell zur Simulation der Schädigungsentwicklung und von Spannungsumlagerungen in ermüdungsbeanspruchten Betondruckzonen numerisch implementiert. Hierfür wurde ein dreistufiges Vorgehen gewählt. Im ersten Schritt wurden experimentelle Untersuchungen konzipiert und durchgeführt. Die Versuche waren eine wichtige Grundlage für das numerische Modell und wurden an zylindrischen und balkenförmigen Probekörpern durchgeführt. Aus statischen und zyklischen Versuchen an den Betonzylindern wurden die Eingangsparameter für das Materialmodell ermittelt und die Balkenversuche dienten zur Validierung des Modells. Diese wurden in einem Resonanzprüfstand durchgeführt, mit dem sich die erforderlichen großen Kräfte mit Belastungsfrequenzen von 19 Hz realisieren ließen. Damit konnten die hohen auftretenden Lastwechselzahlen in kurzer Zeit erreicht werden. Im zweiten Schritt wurde ein additives Dehnungsmodell in den Berechnungsablauf der Finite-Elemente-Software ANSYS Mechanical implementiert. Die zugrundeliegenden Funktionen für die Steifigkeits- und Dehnungsverläufe wurden aus Zylinderversuchen abgeleitet. Im dritten Schritt erfolgte die Validierung der numerischen Implementierung anhand der Balkenversuche. Es wurden unterschiedliche Phänomene betrachtet, wie z. B. die Temperatur- und Dehnungsentwicklung sowie die Veränderungen in der Materialsteifigkeit und die daraus resultierenden Spannungsumlagerungen im Querschnitt. In den experimentellen Untersuchungen der Balkenprobekörper wurde mit der Resonanzprüfmethode ein Ermüdungsversagen durch Abplatzen der Betondruckzone oder durch eine vollständige Zerstörung der Balken erzielt. Die ertragenen Lastwechselzahlen aller Balken lagen deutlich über den Bruchlastwechselzahlen nach fib Model Code 2010. Die Ursache dafür ist eine frühe Schädigung und Steifigkeitsabnahme der am stärksten beanspruchten Bereiche, die dort zu einer nachfolgenden Verringerung der Spannungen führte. In den numerischen Nachrechnungen konnten die in den Balkenversuchen beobachteten Effekte bestätigt werden. Vor allem die am stärksten geschädigten Bereiche konnten mit dem Modell sehr gut abgebildet werden. Es wurde zudem gezeigt, dass sich infolge der Entlastung der geschädigten Bereiche Spannungen in weniger stark beanspruchte Bereiche umlagerten. Dies bestätigte die positive Wirkung der Spannungsumlagerungen auf die Ermüdungslebensdauer der Bauteile. Mit dem Modell können Parameterstudien mit z. B. variierenden Belastungsniveaus oder Vorspannungsgraden durchgeführt werden, um wertvolle Erkenntnisse zur Vordimensionierung ermüdungsbeanspruchter Bauteile und zur Identifikation markanter Stellen in der Geometrie zu gewinnen. Für weiterführende Aussagen ist die Durchführung zusätzlicher Versuche notwendig, um das numerische Modell auf ein breiteres Parameterspektrum zu erweitern. / The fatigue behavior of concrete has so far been studied intensively in the literature using centrically loaded cylindrical specimens. Only a few studies have been carried out on concrete structures subjected to cyclic bending loads. Such structures can be found in wind turbine towers, for example. Research results for high-strength concrete, which is usually used in such structures, are even scarcer. Due to the high number of occurring load cycles and the large cross-sectional dimensions, the testing of the fatigue behavior of such structures is associated with major challenges. Numerical simulations, in contrast, enable the investigation of a large number of parameter variations within a very short period of time, which accelerates the investigation of the fatigue behavior of components. The used material models quickly become very complex and require time-consuming calibration of the input parameters in structural tests. A model for simulating the macroscopic damage processes in concrete subjected to fatigue loading in compressed cross-sections, which can be calibrated using standard cylinder tests, does not yet exist. The present work addresses these aspects and implements an additive strain model to simulate the damage development and stress redistributions in fatigue-loaded concrete compression zones. A three-step approach was chosen for this. In the first step, experimental investigations were designed and carried out. The tests are essential for the numerical model and were carried out on cylindrical and beam-shaped specimens. The input parameters for the material model were determined from static and cyclic tests on the concrete cylinders and the beam tests were used to validate the model. The beam tests were carried out in a resonance-based testing facility, which allowed the realization of the required high forces with load frequencies of 19 Hz. This made it possible to apply the high number of load cycles in a short time. In the second step, the strain model was implemented in the simulation procedure of the finite element software ANSYS Mechanical. The functions for the stiffness and strain development were derived from cylinder tests. In the third step, the numerical implementation was validated using the beam tests. Different phenomena were considered, such as the temperature and strain development as well as the changes in material stiffness and the resulting stress redistributions in the cross-section. In the experimental investigations of the beam specimens, fatigue failure due to spalling of the concrete compression zone or complete destruction of the beams was achieved using the resonance-based testing facility. All beams endured significantly more load cycles than the number of cycles to failure according to fib Model Code 2010. This was due to early damage and a reduction in stiffness in the most heavily loaded regions, which subsequently led to a reduction in stresses. The numerical simulations confirmed the effects observed in the beam tests. In particular, the model was able to simulate the most damaged regions very well. Moreover, it was shown that stresses redistributed to less heavily loaded regions as a result of the relief of the damaged regions. This confirmed the positive effect of stress redistributions on the fatigue life of the structures. The model can be used to carry out parameter studies with varying load levels or degrees of prestressing, for example, in order to gain valuable insights into the predimensioning of fatigue-loaded structures and to identify critical locations in the geometry. For more detailed conclusions, additional tests have to be carried out in order to expand the numerical model to a wider range of parameters.

This thesis describes selection and shows calibration of material model, capable of describing cyclic softening of material. Stress-strain FEM analysis of circumferential blade attachment for last section of rotor blades of steam turbine is performed, expected lifetime of existing attachment is evaluated. Multi-parameter optimization of new-shape attachment was done, resulting in dimensions for new-shape attachment with longer lifetime. Improvements in strain amount in comparison with existing attachment were evaluated and possible RPM increase of turbine with new attachment type was calculated.

Stålbroars approximativa livslängd bestäms av stålets utmattningshållfasthet, då utmattning är en av de främsta anledningarna till att livslängden begränsas. I Sverige existerar ett antal broar som närmar sig slutet av sin livslängd, samtidigt som behovet för kapacitet och kraven på broarna ökar. Flertalet av dessa broar är i behov av upprustning. Däremot är det inte möjligt ur vare sig ett ekonomiskt- eller miljöperspektiv att byta ut alla broar, och därför behöver de broar som är mest kritiska prioriteras. Vid utmattningsdimensionering av stålbroar beaktas hela spänningsvidden, oavsett om spänningarna är i drag eller tryck. En spricka propagerar endast vid dragspänningar, vilket innebär att tryck- spänningar egentligen inte bör vägas in i samband med dimensionering. Detta innebär att en del stålbroar skulle kunna ha en längre livslängd än vad den traditionella dimensioneringen ger. Spänningsintensitetsfaktorn 𝐾 används inom brottmekaniken för att förutspå spänningsintensiteten i närheten av sprickspetsen, och appliceras till linjärelastiska material. Det finita elementprogrammet Abaqus användes när brodetaljen modellerades och analyserades. Brodetaljen representerar en balk med en påsvetsad anslutningsplåt, som utsätts för trafiklasten på en bro och en temperaturlast för att simulera egenspänningar. Detaljen representerar problematiken med utmattning i stålkonstruktionsdelar. Motivet för denna studie är att inga sprickor har hittats under inspektioner av liknande detaljer, det till trots att en del stålbroar teoretiskt sett förbrukat sin livslängd. Studien genomfördes med en mer avancerad modell än vad som vanligtvis skapas för bedömning av utmattning, med syftet att modellera verkligheten mer korrekt. Resultaten visar hur egenspänningarna bidrar till dragspänningar, vilket leder till sprickpropagering i modellen. Vid spricklängden 9,5 mm övergår spänningarna från drag till tryck, och då upphör spricktillväxten. Resultaten visar även att utmattningssprickor kan växa i stålkonstruktionsdelar som i huvudsak utsätts för nominella tryckspänningar, ifall höga egenspänningar uppstår vid anslutningsplåten. / Steel bridges estimated service life is determined by the fatigue strength of the steel, since fatigue is one of the main reasons for limiting the service life. In Sweden there is a number of bridges that approach the end of their service life, while the need of increasing the capacity and demands on bridges grows. The majority of these bridges is in need of reparation. On the other hand, it is not possible either from a financial- or environmental perspective to replace all bridges, and therefore the bridges that are most critical needs priority. In the case of fatigue design calculation of steel bridges, the entire stress range is taken into account, regardless of whether the stresses are in tension or pressure. A crack propagates only at tensile stresses, which means that pressure should not really be considered in the design calculations. This means that some steel bridges could have a longer life span than the traditional design calculation gives. The stress intensity factor K is used within the fracture mechanism to predict the stress intensity near crack tip, and is applied to linear elastic materials. The finite element program Abaqus was used when the bridge detail was modeled and analyzed. The bridge detail represents a beam with a welded connection plate, which is exposed to traffic load at the bridge and a temperature load to simulate residual stresses. The detail represents the problem of fatigue in steel structural parts. The motive for this study is that no cracks have been found during inspections of similar details, despite the fact that some steel bridges theoretically have consumed their longevity. The study is conducted with a more advanced model than usually created for assessment of fatigue, with the purpose of modeling the reality more correctly. The results show how the residual stresses cause tensile stresses, which leads to crack propagation in the model. At a crack length of 9,5 mm, the stresses change from tension to compression, and then the crack growth ceases. The results also indicate that fatigue cracking can grow in steel structural parts that are mainly exposed to compressive nominal stresses, if tensile residual stresses appear at the connection plate.

Hydropower is the most important power balancing resource in the Swedish electrical power system, regulating the power supply to match the load. Consequently, several hydropower plants have periods of stand-by operation where the power production is absent but where several devices within a plant are still active. Such a device is the step-up power transformer, which during stand-by operation still generates no-load energy losses. These losses can accumulate to a considerable amount of energy and costs during the long technical lifetime of the apparatus. One option to minimize these no-load energy losses is by turning the transformer off when its generating unit is in stand-by operation. However, when this transformer operational change has been explained to experts in the field, the most common response has been that a more frequent reenergizing of a transformer leads to higher risks for errors or transformer breakdowns. This study aimed to analytically investigate three effects from this operational change. First, the potential of fatigue failure for the windings due to the increased sequences of inrush current. Secondly, the thermal cycling as a consequence of change in present losses. Lastly, the energy and economic saving potentials for hydropower plants where this operational adjustment is applied. The study used both established as well as analytical tools explicitly created for this study. These were then applied on currently active transformers in different plant categories in Fortum’s hydropower fleet.  The study primarily showed three things. Firstly, risk of fatigue failure due to the increased presence of inrush currents did not affect the transformer’s technical lifetime. Secondly, the thermal cycling changes were slightly larger with absent no-load losses during stand-by operation. The average temperature for the transformer decreased, which in general is seen as a positive indicator for a longer insulation lifetime and thus the transformer’s technical lifetime. Finally, the created frameworks showed the potential of saving energy and money for all plant categories, where the potential grew with the installed production capacity and the stand-by operation timeshare. Despite the simplifications made to describe the complex reality of a transformer operating in a hydropower plant, this thesis contributes to lay a foundation for future investigation of an easy adjustment to avoid unnecessary energy losses and costs for transformers in hydropower plants.

[pt] Neste trabalho, o fator de gradiente de tensões elastoplástico na frente da raiz de entalhe é utilizado para investigar o efeito real do entalhe na resistência à fadiga, quantificado pelo fator de concentração de tensões à fadiga. Este é geralmente menor que o fator de concentração de tensões do entalhe, o parâmetro linear elástico (Fator de concentração de tensão) , devido à tolerância do material a trincas curtas não propagantes. Considerando que a plasticidade localizada na vizinhança da raiz do entalhe afeta o comportamento de crescimento de trincas curtas dentro da zona plástica induzida pelo entalhe, uma abordagem baseada na Mecânica da Fratura é proposta para abranger os efeitos dos campos de tensões e deformações elastoplásticas no cálculo do (Fator de concentração de tensão de fadiga). Análises bidimensionais por elementos finitos são adotadas para calcular fatores de intensidade de tensão de espécimes planos e entalhados. O modelo de encruamento de Ramberg-Osgood e a regra de Neuber são usados para obter aproximações de fatores de intensidade baseados em deformação. Para validação da metodologia, as previsões numéricas geradas são comparadas com dados experimentais de S-N coletadas da literatura para espécimes com (Fator de concentração de tensão) . Estes possuem furo circular central ou entalhes tipo U ou V, são feitos de diferentes materiais e testados sob cargas axiais com razão igual a -1, 0 ou 0,1. A comparação mostra boa concordância e prova que a solução elastoplástica proporcione maior precisão do que a linear elástica. Os resultados mais discrepantes são obtidos em razões de carga de 0 e 0,1, no entanto, eles podem ser significativamente melhorados quando é considerada a correção de efeitos de tensão média não nula. / [en] Elasto-plastic stress gradient factors ahead of notch tips are used to evaluate actual notch effects in fatigue strength, quantified by fatigue stress concentration factor (Fatigue stress concentration factor). Usually, it is smaller than the linear elastic stress concentration factor of the notch, (Stress concentration factor) , due to material tolerance to non-propagating short cracks. Considering that local plasticity around notch tips plays a significant role in the growth behavior of short cracks within the notch plastic zone, a sound mechanical methodology is proposed to account for the effects of elasto-plastic stress and strain fields in the actual (Fatigue stress concentration factor) value. Two-dimensional finite element analyses are conducted to compute stress intensity factors of smooth and notched specimens. Ramberg-Osgood model and Neuber s rule are used to achieve approximations for strain-based intensity factors. For methodology validation, numerical predictions are compared to experimental stress-life data of center, U, and V-notched plate specimens made of different materials and tested under uniaxial load ratios of −1, 0, and 0.1 collected from the literature. The comparisons show good agreement proving that the elasto-plastic solution provides more accuracy than the linear elastic one. The most discrepant results are obtained at load ratios of 0 and 0.1, and they can be significantly improved if non-zero mean stress effects are considered.

Hydropower is the most important power balancing resource in the Swedish electrical power system, regulating the power supply to match the load. Consequently, several hydropower plants have periods of stand-by operation where the power production is absent but where several devices within a plant are still active. Such a device is the step-up power transformer, which during stand-by operation still generates no-load energy losses. These losses can accumulate to a considerable amount of energy and costs during the long technical lifetime of the apparatus. One option to minimize these no-load energy losses is by turning the transformer off when its generating unit is in stand-by operation. However, when this transformer operational change has been explained to experts in the field, the most common response has been that a more frequent reenergizing of a transformer leads to higher risks for errors or transformer breakdowns. This study aimed to analytically investigate three effects from this operational change. First, the potential of fatigue failure for the windings due to the increased sequences of inrush current. Secondly, the thermal cycling as a consequence of change in present losses. Lastly, the energy and economic saving potentials for hydropower plants where this operational adjustment is applied. The study used both established as well as analytical tools explicitly created for this study. These were then applied on currently active transformers in different plant categories in Fortum’s hydropower fleet. The study primarily showed three things. Firstly, risk of fatigue failure due to the increased presence of inrush currents did not affect the transformer’s technical lifetime. Secondly, the thermal cycling changes were slightly larger with absent no-load losses during stand-by operation. The average temperature for the transformer decreased, which in general is seen as a positive indicator for a longer insulation lifetime and thus the transformer’s technical lifetime. Finally, the created frameworks showed the potential of saving energy and money for all plant categories, where the potential grew with the installed production capacity and the stand-by operation timeshare. Despite the simplifications made to describe the complex reality of a transformer operating in a hydropower plant, this thesis contributes to lay a foundation for future investigation of an easy adjustment to avoid unnecessary energy losses and costs for transformers in hydropower plants