No description available.

No description available.

No description available.

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第18653号 / 工博第3962号 / 新制||工||1610(附属図書館) / 31567 / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 杉浦 邦征, 教授 白土 博通, 教授 河野 広隆 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Infrastrukturen i Sverige åldras, uppskattningsvis finns det drygt 2000 broar i landet som är 70 år eller äldre. Det finns flera aspekter att titta på för att utvärdera äldre broars kondition och FE modellering är ett vanligt verktyg som används vid utvärdering. För stålbroar är det ofta utmattning som sätter gränsen för hur länge de kan hållas i drift under säkra förhållanden. Syftet med arbetet var att utvärdera nyttan av FE modeller med olika precision med hänsyn till noggrannheten i deras genererade resultat och kostnad. En stålbro har studerats i detalj, Gamla Lidingöbron och specifikt två punkter på dess bågspann. Ritningar av bron och mätdata från forskningsprojektet \textit{Smart tillståndsbedömning, övervakning och förvaltning av kritiska broar} har legat till grund för arbetet. Mätdata utgjordes av tidshistorier över spänningsvariationer i de två studerade punkterna vid flertalet tågpassager över bron. Punkterna finns på bågspannets sekundära bärverk. Fyra modeller har skapats i BRIGADE/Plus med olika precisionsgrad och spänningshistorier för de studerade punkterna har genererats. Spänningshistorierna har sedan utvärderas med Palmgren-Miners delskadeteori för utmattning. Den förväntade ekonomiska nyttan för varje modell har uppskattats beroende av analyskostnad, sannolikheten för utmattningsbrott samt kostnad för ett eventuellt brott. På grund av brons strukturella verkningssätt blev de förenklade modellernas utmattningsresultat mycket lika resultaten från modellen med hög precisionsgrad. De verkliga axellasterna var mindre än de dimensionerande som användes i modellerna. Detta ledde till att de uppkomna spänningarna i stålet från modellerna blev större än de verkliga, men visade ett liknande beteende. Resultatet blev att modellernas utmattningskapacitet enbart utgjorde ca en femtedel av kapaciteten enligt mätningarna. Resultatet från beräkningen av den förväntade nyttan visade att det inte är ekonomiskt motiverat att använda en modell med hög precision framför en eller flera förenklade modeller. Slutsatsen blev att hög precision i en teoretisk modell inte entydigt är bättre än en förenklad modell. Arbetet har utförts vid institutionen för bro och stålbyggnad på KTH i samarbete med Sweco Civil AB. / The infrastructure in Sweden is aging. More than 2 000 bridges in the country are 70 years or older. When assessing the condition of older bridges, several aspects should be taken into account. FE modeling is one common tool to use in a bridge assessment. Fatigue is generally what limits the service life of steel bridges. The aim of this work was to evaluate the utility of higher precision in FE models, regarding the accuracy of their generated results and costs. One steel bridge has been studied in detail, Old Lidingö Bridge and specifically two points on its arc span. Drawings of the bridge and measurement data from the research project \textit{Smart Condition Assessment, Surveillance and Management of Critical Bridges} has provided the basis of this work. Measurement data has been collected from the two selected points on the bridge with strain gauges, registering the time history of the variations in tension during train passages on the bridge. Four models with different levels of precision have been created in BRIGADE/Plus. These models have generated time histories of the varying tension during  train passages. The time histories have been evaluated with Palmgren-Miner's cumulative damage model for fatigue. The expected economic utility of each set of models has been estimated based on the cost of the analysis, the likelihood failure caused by fatigue and the cost of failure. Because of the structural behavior of the bridge, the results of the simplified models became very similar to the results of the high precision model. The actual axle loads were less than the design loads used in the models. Because of this, the calculated tensions in the models became larger than the actual tensions. This also resulted in the fatigue capacity of the models only being about one fifth of the capacity according to the measurements. The calculation of the expected utility showed that it is not economically justified to use a model with higher precision over models with less precision. High precision in a theoretical model is not unambiguously better than a simplified model. The work has been carried out at the Department of Structural engineering and bridges at KTH in cooperation with Sweco Civil AB.

No description available.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第23483号 / 工博第4895号 / 新制||工||1765(附属図書館) / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 杉浦 邦征, 教授 清野 純史, 教授 八木 知己 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Les tôles courbes sont de plus en plus fréquemment utilisées dans le domaine des ouvrages d'art à la fois pour raisons esthétiques et pour leurs performances mécaniques. Leur dimensionnement est cependant difficile du fait de l'absence de méthode dédiée dans les normes européennes, en particulier en ce qui concerne le voilement de l'Eurocode 3. Le recours à la modélisation par éléments finis est donc nécessaire mais bien souvent coûteux. La présente étude vise ainsi à développer une méthode simplifiée pour le dimensionnement des panneaux cylindriques soumis à une compression uniaxiale. La méthodologie utilisée est basée sur la démarche générale de l'Eurocode 3 pour tous les types de vérifications de stabilité. L'expression analytique obtenue est en bon accord avec les standards des recommandations de l'Eurocode. Cependant, cette démarche peut être difficile à reproduire dans la pratique quotidienne des ingénieurs face à un nouveau problème tel qu'une combinaison de chargements par exemple. Une méthode alternative basée sur la méthode des plans d'expérience numérique a donc été développée. Basée sur des outils statistiques, l'application de cette méthode permet facilement d'obtenir un résultat avec une bonne précision tout en nécessitant que peu simulations / The use of curved panels for the construction of steel bridges has become more and more popular due to structural performance, traffic and aesthetic requirements. Their design is however made difficult due to a lack of specifications, especially in European Standards. The use of finite element modeling is often necessary. This however requires very sophisticated and expensive models in terme of calculation times. The present study aims thus at developing a method for predicting the ultimate strength of cylindrical curved panels subjected to uniform axial compression. This method is based on the formal procedure recommended by Eurocode 3 for all types of stability verifications. The agreement with the numerical results is thus very good. However, this approach seems complex to reproduce in practice user when confronting a new problem such a combinaison of loading. An alternative method based on design of computer experiment is thus developed. Based on statistic notions, this turnkey methodology provides good results with the fewest simulations possible

<p dir="ltr">Girder ends of steel bridges can be corrosion damaged due to deicing salts, water, and other contaminants leaking from deck expansion joints. When the corrosion becomes significant, it can decrease the sectional properties of end steel girders and eventually reduce structural resistance against bearing and shear. Conventional methods that are typically used to repair corrosion-damaged girders require a substantial amount of time and resources to complete and often cause public inconvenience due to traffic lane closure. Therefore, there is a need for practical, rapid, and robust repair methods that can be implemented by maintenance personnel of a local Department of Transportation (DOT).</p><p dir="ltr">In this study, five innovative repair methods for corroded steel girders were evaluated through a selection process called the House of Quality Matrix, a commonly used tool in the consumer product industry. After completing the evaluation and additional numerical simulations, the "Sandwich Panel" repair method was selected for further investigation. The main concept of the proposed "Sandwich Panel" repair method is the encasement of the corroded region with a filler material reinforced by threaded rods. Two thin steel plates installed on both girder sides serve as stay-in-place formwork, expediting the installation process. This repair method eliminates labor-intensive steps of jacking, welding, and formwork disassembly, making it more cost-effective and less time-consuming.</p><p dir="ltr">The structural performance of the method was evaluated experimentally by conducting seven large-scale tests. Various test parameters were considered in the tests, including i) threaded rod layout, ii) filler material, and iii) support condition. The test specimens were corrosion-damaged steel girders from decommissioned highway bridges in Indiana. The experimental results indicate that the method is effective enough to recover the girder's original design strength. The experimental evaluation was followed by a numerical parametric study using finite element models benchmarked using the experimental results. Detailed design guidelines and recommendations were developed based on the experimental and numerical results.</p>

No description available.