İstinat duvarlarına etkiyen dinamik toprak basınçlarına yerel zemin sınıfı ve spektral ivme katsayısının etkisi

Abstract

Tarih boyunca meydana gelen yıkıcı depremler, yalnızca üstyapı sistemlerinde değil, altyapı sistemleri ve destek yapılarında da ciddi hasarlara ve yıkımlara neden olmuştur. Bu yapılardan biri olan istinat duvarları, eğimli arazilerde toprak kütlesinin hareketini engellemek, şev stabilitesini sağlamak ve yakındaki yapıların güvenliğini korumak amacıyla inşa edilen mühendislik yapılarıdır. Özellikle yol, köprü, bina gibi önemli yapıların güvenliğini sağlayan istinat duvarları bulundukları konum itibarıyla bu yapılarda oluşabilecek hasar boyutunu doğrudan etkileyebilmektedir. Deprem etkisi altındaki bölgelerde bu yapıların güvenliğinin sağlanması geoteknik mühendisliğinin önemli konularından biridir. Bu doğrultuda, deprem bölgelerinde yer alan istinat duvarlarının tasarımı yapılırken deprem yüklerinin doğru bir şekilde hesaplara dahil edilmesi gerekmektedir. Çalışma kapsamında öncelikle istinat duvarına etki eden dinamik toprak basınçlarının belirlenmesinde kullanılan Mononobe Okabe, Seed&Whitman yöntemi ile Türkiye Bina Deprem Yönetmeliği 2018'de tanımlanan hususlar doğrultusunda dinamik toprak basınçları hesaplanarak çeşitli parametreler altında aralarındaki farklar grafikler yardımıyla incelenmiştir. Tez çalışmasının temel amacı, istinat duvarlarına etki eden dinamik toprak basınçlarına yerel zemin sınıflarının ve kısa periyot spektral ivme katsayısının etkisini incelemektir. İstinat duvarının dinamik tasarımı yapılırken yalnızca deprem yüklerinin değil, aynı zamanda zemin özelliklerinin etkisinin de detaylı bir şekilde incelenmesi gerekmektedir. TBDY-2018 sismik etkilerle birlikte yerel zemin sınıflarını dikkate aldığı için güncel mühendislik standartlarıyla uyumlu, daha kapsamlı ve gelişmiş bir metodolojiyi ortaya koymaktadır. Bu doğrultuda, TBDY-2018 kapsamında yer alan istinat duvarı tasarım hususları dikkate alınarak farklı yerel zemin sınıfları (ZB, ZC, ZD) ve spektral ivme katsayıları (Ss) altında, yayılı yük etkisindeki istinat duvarlarına etki eden toplam toprak basıncı (Pt) ve devirici momentler (Edev) hesaplanmıştır. Hesaplamalar farklı zemin özellikleri ve yükleme durumları için tekrarlanarak bu değişkenlerin toprak basınçları üzerindeki etkileri incelenmiştir. Farklı yükleme koşulları altında elde edilen tasarıma esas teşkil edecek statik ve dinamik toprak basınçlarından oluşan toplam aktif itki Pt ve devirici moment Edev değerlerini, geoteknik mühendislik uygulamalarında pratik amaçlar doğrultusunda seçilen değişkenlere bağlı olarak tahmin edebilmek amacıyla Regresyon analizi ile bağıntılar geliştirilmiştir. Elde edilen sonuçların, mühendislik uygulamalarında karşılaşılabilecek farklı zemin ve yükleme koşulları için güvenli, pratik ve hızlı çözümler geliştirilmesine katkı sağlanması amaçlanmaktadır.

Throughout history, destructive earthquakes have caused not only extensive damage to superstructures but also significant destruction in infrastructure systems and supporting structures. Among these, retaining walls play a critical role in maintaining slope stability, preventing soil movement, and ensuring the safety of nearby structures such as roads, bridges, and buildings. Positioned at strategic locations, retaining walls significantly influence the extent of damage that may occur during seismic events. Therefore, in earthquake zones, ensuring the seismic safety of these walls is a crucial issue in the field of geotechnical engineering. An accurate definition and incorporation of earthquake loads into design calculations is essential for the safe performance of retaining structures under dynamic conditions. This thesis focuses on the evaluation of static and dynamic earth pressure methods and investigates the influence of local soil class and short-period spectral acceleration coefficient (Ss) on dynamic earth pressures acting on retaining walls. The main objective of this study is to evaluate how the dynamic earth pressure on retaining walls is affected by the local soil classes (ZB, ZC, ZD) and the short-period spectral acceleration coefficient defined in TBDY-2018. To achieve this, the influence of additional variables such as wall height (H), internal friction angle of the backfill soil (ϕ), magnitude of the surcharge load (q) and slope angle of the backfill (β) were also systematically investigated. These variables were chosen because of their strong influence on the seismic earth pressure behavior. The backfill soil is assumed to be coarse-grained and the distributed load is assumed to start directly from the backfill soil. In the literature, there are numerous studies investigating the behavior of dynamic earth pressures acting on retaining walls. Comparative analyses between national and international seismic codes have revealed that previous codes require improvements, especially considering different wall types and soil classifications. Studies conducted under both DBYBHY-2007 and its revised version TBDY-2018 have shown that the updated code introduces higher static equivalent seismic coefficients, resulting in increased earth pressures. In line with these studies, this thesis investigates the effects of local soil class and spectral acceleration coefficient on dynamic earth pressures, performing analyses under varying soil properties, surcharge loads, and wall heights in accordance with TBDY-2018, and evaluating their influence through graphical comparisons. The study initially presents calculations for the Mononobe-Okabe method, the Seed & Whitman approach and the considerations defined in the Turkish Building Earthquake Code 2018 (TBDY-2018) to calculate dynamic earth pressures. These three approaches are comparatively analyzed under various geotechnical and seismic parameters and the differences are evaluated. The results are visualized with tables and graphs and the variation of earth pressures under various parameters for each method is examined. In the second and main phase of the study, dynamic earth pressures and overturning moments were calculated according to TBDY-2018 for a comprehensive range of variable combinations. A total of approximately 10.000 scenarios were analyzed, covering multiple values of each variable: wall height H=4, 5, 6, 7, 8 m, internal friction angle ϕ=25.0˚, 27.5˚, 30.0˚, 32.5˚, 37.5˚, 40.0˚, surcharge load q=10, 20, 30 kPa, and backfill slope β=5˚, 10˚, 15˚, 20˚. These calculations were carried out separately for each of the local soil classes (ZB, ZC, ZD) and for varying Ss values (0.25 to 1.50), allowing a parametric investigation of each input on total earth pressure (Pt) and total overturning moment (Edev). The results demonstrated that local soil class and Ss are among the most influential parameters affecting dynamic earth pressures. The increase in total earth pressure (Pt) and total overturning moment (Edev) was more significant in flexible soil classes, particularly ZC, as Ss increased. This highlights the amplifying effect of more deformable soils under seismic loading conditions. Additionally, the internal friction angle (ϕ) showed a notable effect on reducing Pt at low seismic intensities; however, this effect diminished as Ss increased, indicating the dominance of inertial forces under strong earthquakes of the wall. The data obtained in this study are presented in tabular and graphical form to facilitate practical interpretation and application. The effect of each variable is analyzed separately to understand how changes in wall geometry or seismic loading affect the overall design forces acting on the wall. As a further step in the study, a multivariate linear regression analysis was conducted to develop predictive equations for estimating the Pt and Edev based on selected input parameters. The purpose of this analysis was to provide simplified, practical tools for engineers to estimate seismic design forces without the need for detailed iterative calculations. The regression models were derived by using the results obtained from the parametric analysis performed under the framework of TBDY-2018. Independent variables included the logarithmic transformations of parameters such as the internal friction angle of the backfill soil (ϕ), the short-period spectral acceleration coefficient (Ss), and the surcharge load (q), while the dependent variables were the logarithmic values of Pt and Edev. For each local soil class, separate regression models were developed to reflect the soil-specific behavior under seismic loading. Furthermore, in cases where spectral acceleration values were relatively high (Ss ≥ 1.25), deviations were observed due to the nonlinear amplification effects in flexible soils; therefore, separate regression models were also constructed for higher seismic scenarios to increase accuracy. The multivariate linear regression models developed based on the data obtained in this study are intended to enable the rapid and practical estimation of total earth pressure and overturning moment values in engineering applications. In conclusion, this thesis provides a detailed and comprehensive evaluation of the seismic behavior of retaining walls, particularly focusing on the effects of local soil conditions and site-specific seismic parameters. The comparative analysis of the three different calculation methods provides valuable insight into the reliability and limitations of each approach. The findings underscore the importance of incorporating local soil classification and spectral acceleration data into retaining wall design, especially in regions with high seismic risk. The results of this study are expected to contribute to the development of practical, efficient, and safe design solutions that are compatible with varying soil and loading conditions encountered in the field. Furthermore, the methodological framework and findings can serve as a foundation for future numerical and experimental studies aimed at enhancing seismic design standards for retaining walls in geotechnical engineering.