Öngerilmeli Betonarme Köprülerde Zamana Bağlı Yerdeğiştirmeler

Abstract

Bu çalışma kapsamında 7 tanesi Türkiye’nin çeşitli bölgelerinde (Şehzadeler, Kömürhan, Gülburnu, Beylerderesi, Akarsın, İmrahor, Tigem), bir tanesi de Palau Cumhuriyeti’nde olmak üzere 8 tane kutu kesitli, öngerilmeli köprünün zamana bağlı yerdeğiştirmeleri incelenmiştir. Uygulamada genelde ilk birkaç yılda oluşan yer değiştirmelerin zamanla azalarak duracağı varsayılır. Gerçekte ise yerdeğiştirmeler yıllar sonra da artmaya devam eder. Bu yer değiştirmelerin doğru olarak bulunması yapının kullanım süresini önemli ölçüde etkilemektedir. Uzun süreli kullanımda olması düşünülen öngerilmeli yapılarda zamana bağlı yerdeğiştirmeleri oluşturan en önemli etkiler sünme ile büzülmedir. Sünme, büzülme etkileri hem malzemeye hem de yapının yüklenme durumuna bağlı olarak değişmektedir. Öngerilmeli yapıların enkesitleri, yapının kullanım süresi boyunca öngerilme ile sabit bir basınç basınç etkisi altında kalmaktadır. Bu sabit basınç etkisi, yapıda sünme etkisi oluşturarak zamana bağlı yerdeğiştirmelere neden olur. Bunun yanında yapının bulunduğu ortam koşulları nedeniyle büzülme etkileri de yapının kullanım süresi boyunca önemli ölçüde etkindir. Kullanımda olan sünme ile büzülme modelleri çevre koşulları, köprü yapım süresi, kesit özellikleri, beton dayanımı, beton karışım oranı gibi özellikler göz önünde bulundurularak geliştirilmiştir. Tez kapsamında bu modellerden Bazant tarafından geliştirilen B3 Modeli, ACI 209R-92 Modeli, Eurocode 2 Modeli ile TS3233 Modeli kullanılarak karşılaştırmalı bir çalışma yapılmıştır. Çalışmanın sonunda da kullanılan modellerin verdiği değişik sonuçların nereden kaynaklandığı tartışılmıştır. Bu çalışmada öncelikle kullanılan dört farklı yöntem tanıtılmış, daha sonra da bu yöntemler bütün köprülere ayrı ayrı uygulanmıştır. Kullanılan güncel yöntemler arasında en büyük yer değiştirmeleri Bazant’ın B3 Modeli vermiştir. B3 Model ile diğer modellerin verdiği sonuçlar arasındaki bu farklılığın ise büyük ölçüde B3 Model’in sünme, büzülme etkilerini daha gerçekçi göz önüne almasından kaynaklandığı sonucuna varılmıştır. Eurocode 2 ile ACI 209R-92 Modellerinin verdiği sonuçlar ise birbirine yakın çıkmıştır. Bu modellerden Eurocode 2’nin sünme etkisini, ACI 209R-92 Modeli’nin ise büzülme etkisini daha az öngördüğü görülmüştür. TS3233’ün sünme ile büzülme yaklaşımı incelendiğinde ise bu yönetmeliğin oldukça eski (1979) olması nedeniyle güncel yöntemlerin soruna ayrıntılı yaklaşımı karşısında güncelliğini yitirdiği anlaşılmıştır. Ülkemizde öngerilmeli yapılar için kullanılan TS3233 yönetmeliği sünme ile büzülme etkilerini belirli katsayılar aracılığıyla göz önüne almaktadır. Bu katsayıların verildiği ilgili çizelgeler ise yalnızca 3 yıla kadar düzenlenmiştir. Bu nedenle sonuç olarak bu yönetmeliğin güncellenmesi gerektiği kanısına varılmıştır. Zamana bağlı yer değiştirmeleri bulunan köprüler arasında en büyük yer değiştirmenin Tigem Köprüsü’nde olduğu görülmüştür. Bu köprü, diğer 7 köprü gibi dengeli konsol yöntemi ile yapılmamıştır. Tigem Köprüsü’nde bulunan zamana bağlı yer değiştirmelerin diğer köprülere göre daha fazla çıkmasının nedeninin de bu yapım farklılığından kaynaklandığı sonucuna varılmıştır. Bu yüzden kullanılan sünme ile büzülme modellerinin köprülerin yapım yöntemiyle doğrudan ilgili olduğu görülmüştür.

In this study, time dependent displacements of box-sectioned concrete bridges, which are Palau, Beylerderesi, Şehzadeler, Akarsın, Gülburnu, Kömürhan, Imrahor and Tigem, were determined. Several models have been proposed for predicting shrinkage and creep for the purpose of helping to the designers to estimate the creep and shrinkage strain more reliably. These models aim to estimate long term effects for a given structure relatively quick with sufficient precision by using known parameters. Scope of this study contains estimation of time-dependent effects by B3, ACI 209R-92, EC 2 Models for 10000 days and TS 3233 Model for 1000 days. Concrete experiences volume changes throughout its service life. When loaded, concrete experiences an instantaneous recoverable elastic deformation and a slow inelastic deformation called creep. Creep of concrete is composed of two components, basic creep, or deformation under load without moisture loss and drying creep, or deformation under drying conditions only. Deformation of concrete in the absence of applied load is often called shrinkage. Due to the these effects, the pre-stress in tendons get reduced with time which should be taken into consideration very carefully. Almost all large spanned, pre-stressed segmental box girders are deflected excessively by the time. Main problem of these excessive deflections is related with designing procedure. In use practices do not apply realistic models for creep and shrinkage which cause unexpected deflections over the time. To predict the strength and serviceability of reinforced and pre-stressed concrete structures, structural engineer requires an appropriate description of the mechanical properties of the materials, including the prediction of the time-dependant strains of the hardened concrete. The prediction of shrinkage and creep is important to assess the risk of concrete cracking, and deflections due to stripping-reshoring. In pre-stressed concrete structures, dead loads would cause downward deflections and in contrary pre-stressing would induce upward deflections. Difference between these counter effects is considerably small at first but time-dependent effects increase this difference significantly. Therefore realistic prediction of displacements due to creep and shrinkage is quite important. Pre-stressed bridge codes consider creep and shrinkage factors as constant values throughout box-sectioned spans. Designs made of according to in use codes would fail because of excessive time-dependent displacements (Palau, Kömürhan). For this reason it is thought that Bazant’s B3 Model, which takes environmental effects, construction time, cross-sectional properties, concrete strength and mixture into account more realistically, is the most advanced approach to ensure a safer design The current empirical creep prediction models like ACI 209R-92, TS 3233 and EC 2 lead to underestimation of creep deflections. These current codes give an incorrect shape of long-time creep curves and incorrectly the existence of final creep comparing to Bazant’s B3 model that is originally calibrated by a worldwide laboratory database. Main objective of this thesis is to show difference in design codes and Bazant’s B3 model. Bazant’s B3 model takes into account more factors than other models does and it is more complex than design code models. It has a different structure than the other were it enables the calculations of separate compliance functions for the basic creep and drying creep. The Model B3 uses the following factors to predict concrete creep; aggregate to cement ratio, cement content, cement type, water to cement ratio, water content, age at loading, age of sample, applied stress, cross-section shape, curing conditions, compressive strength at 28 days, duration of loading, effective thickness, elastic modulus at loading, elastic modulus at 28 days, relative humidity, and time of drying starts. The ACI 209R-92 code uses the following factors to predict creep and shrinkage: the age at loading, ambient relative humidity, the surface to volume ratio, slump of fresh concrete, fine aggregate percentage, cement content, air content and curing period of the concrete. The ACI 209R-92 takes these aspects into account using correction factors which are applied to measured or suggested values. In the absence of needed data, required parameters are given as constant values. For this model both suggested and calculated parameters are compared to each other. Most code equations for the EC 2 attributes creep and shrinkage of concrete to the relative humidity, dimension of the element, and composition of the concrete. It also states that creep is influenced by the maturity of the concrete when the load is first applied and depends on duration and magnitude of the sustained load. TS 3233 is the code that regulates the construction of pre-stressed structures in Turkey. This code does not have a detailed creep and shrinkage approach. Required parameters are mostly tabulated for certain environmental and concrete properties in this model. So calculations were mostly conducted with the values interpolated from that given tables. TS 3233 Model assumes that creep and shrinkage remain constant after 3 years. Therefore calculations of the model are limited with 1000 days instead of 10000 days. Required parameters of aforementioned models require specific concrete mixture, environmental conditions, age of concrete at loading, age of concrete when drying starts and cross-sectional properties of the bridges. For the concrete mixture, composition of C 35 and C 40 concretes were taken from the laboratory of 15th Branch of General Directorate of State Hydraulic Works (DSİ 15.Şube Müdürlüğü). Environmental conditions were checked from the General Directorate of Meteorology (Meteoroloji Genel Müdürlüğü). Cross-sectional properties are used as the mean of the variable cross-sections. Since cross-sections of the bridges vary throughout the span, effective thickness is taken as the arithmetic mean of the cross-sections. Age of concrete at loading and age of concrete when drying starts are also used as the mean values. All the bridges except Tigem is constructed with the free cantilevering method. That’s why each segment of those bridges is considered to gain their strength in 7 days. In order to find the value of loading time in a consistent way, arithmetic mean of the construction time is taken. First part of the thesis contains literature review of the creep and shrinkage studies until today. Then the introduction of the aforementioned bridges are given in the second part. In the third part, explanations and approaches of the models to predict the long-term displacements were given explicitly and for an example, the models are applied to Şehzadeler Bridge step by step. In the fourth part, models are applied to the other bridges. In this part, explicit calculations are not conducted again, instead related values are given in tables. Finally, the fifth part of the study discusses the differences in results from different models. Application of the models yielded that B3 Model gives the maximum strain comparing to other models at the end of 10000 days. It is concluded that this outcome is directly related with the more sensitive approach of the mentioned model which could be exemplified in the B3 Model’s investigation of the creep as basic and drying creep to state the interconnected nature of creep and shrinkage. Results of ACI 209R-92 and EC 2 Models are found relatively similar to each other. Comparion of these two models showed that even if the total results are similar, ACI 209R-92 finds the shrinkage whereas EC 2 finds the creep less comparing to each other. As it mentioned before TS 3233 Model is limited with 3 years. This restriction is not compatible with the other models because all the other methods suggest that creep and shrinkage effects increase throughout the age of structure. Thus it is concluded that TS 3233 does not provide sufficient information anymore and should be restored to follow more up to date design approaches. Comparison of 8 different bridges yielded that maximum time-dependent deflection due to creep and shrinkage occurs at Tigem Bridge. This bridge was not constructed with the free cantivelering method like others. So, in general it is concluded that construction method is directly related with the time-dependent losses of the pre-stressed structures.