Kohezyonsuz zeminlerde inşa edilen tünellerde zemin yüzeyindeki oturmalara etki eden faktörlerin model deneyleri ile incelenmesi

Abstract

ÖZET Günümüzde metro sistemi şehir içi ulaşımında, tünel ise demiryol ve yol inşaatında geniş olarak yer almaktadır.Bu nedenlerle tünel inşa yöntemlerinde çeşitli gelişmeler yapılmakta, geniş araştırmalar yer almaktadır. Metro in - şaatlarında, özellikle oturmaya hassas bina bölgelerinde zemin üst yüzü düşey şekil değiştirmeleri ve oluşan eğil meler önemli etkenler olmaktadır. Zemin yüzündeki en büyük şekil değiştirme değeri ve otur maların tünel ekseninden itibaren dağılımının önceden be lirlenmesi tasarım safhasında önem kazanmaktadır. Zemin yüzündeki oturmalara ek olarak tünelin kendisi ve yakın yeraltı yapıları tünel civarındaki zeminde meydana gelen de formasyonlardan etkilenmekte dolayısıyle kütle hareket leri önem kazanmaktadır. Bu çalışma, tünel enkesit şeklinin, genişliğinin ve tünel üstü zemin yüksekliğinin, inşa olunacak tünelin oturması na bağlı olarak zemin üst yüzünde meydana gelen oturmala rı ne derecede etkilediğini belirlemek amacıyla gerçek - leştirilmiştir. Yarım daire, elips, atnalı ve dikdörtgen enkesitli model tünellerle yapılan deneylerde tünel ge nişliği B ve tünel üstü zemin yüksekliğinin tünel geniş - ligine oranı h/B değişken olarak alınmıştır. Model malze mesi olarak, genellikle yüzeye yakın yeralan zeminleri simüle etmek amacıyla gevşek sıkılıktaki kum kullanılmış tır. Sonuç olarak, model deneylerinde uygulanan bütün enkesit- lerde zemin yüzündeki en büyük oturma değerinin tünel en kesit genişliğine bağlı olarak doğrusal arttığı belirlen miştir. Oturma değerlerinin genel olarak h/B < 2 ye kadar doğru orantılı artış gösterdiği h/B >2 olan değerlerde ise bu değerlerin değişip azaldığı, h/B=3 ten itibaren de ade ta sabit kaldığı saptanmıştır. Tünel enkesit tiplerinin bilhassa h/B < 2 ye kadar etkili olmadığı, h/B > 3 den itibaren elips enkesitin diğerlerin den daha iyi olduğu saptanmıştır. xıv

SUMMARY MODEL TESTS TO DETERMINE THE FACTORS AFFECTING THE SETTLEMENTS ABOVE TUNNELS IN COHESIONLESS SOILS The rapid industrialisation together with social and urban development brings the need for construction o f underground structures. A great majority of these structures are the tunnels which are built for the transportation purposes or conveyance of materials such as water and sewage. Tunnels constructed in urban areas are usually located at shallow depths for functional and economic reasons and usually at shallow depths soft soils take place. The state of the art of tunnelling in soft ground has undergone substantial change during the last decades. Tunnels had been built nearly only by primitive methods 30 years ago. With the widespread adoption of excavation facilities, soil tunnelling preferred with increasing amount. The advantage of tunnelling when compared with cut-and- cover is the decreased amount of the influence on utilities and on the flow of traffic and the underpinning of neighboring structures. In relation with these developments, the shield method that has been used in the past for several decades has been so developed that it is possible to use this method today in every soft ground. In addition to shield method New Austrian Tunnelling Method has also been used successfully in shallow soft ground tunnels. The developments in tunnelling has been used especially in the metro constructions. One of the major problems associated with tunnelling in urban areas is settlements. Settlements can be a problem with soft ground tunnelling in urban areas where buildings, both modern and ancient,, can be put at risk; services too, can be endangered and it has been deemed necessary to divert services in some cases. xvAs a result, reliable prediction of the extent and amount of the settlements and ground movements caused by tunnelling is required. New Austrian Tunnelling Method has been widely used since the beginning of 50 's. The monitoring of ground movements during tunnel construction is now performed as a matter of course of many projects. The simplest monitoring consists of the measurement of surface settlements. The monitoring of deep settlement points is also often performed. The data from field observations and measurements are used to provide simple analytical tools that enable better prediction of the amount of settlements and ground movements caused by tunnelling through soft ground to be made. In the light of these developments in the measuring technique, it has been found necessary to place the instrumentation before the beginning of the tunnel construction. After examination of ground displacements measured by investigators, it has been observed that a great part of the total deformations takes place at the opening and before the permanant lining. The long term displacements of tunnels in soft ground are smaller than the displacements occurred during the construction. As a result, tunnels with temporary lining in soft ground and deformations at the construction stage have been considered important in tunnel engineering. Although tunnels in soft ground are being widely used in tunnel projects, there is very little understanding of the mechanisms involved. In the past, theoretical studies have been made to determine the settlements at the surface and around the tunnel. Prediction and control of ground movements due to tunnelling has been approached by empirical rules based on field observations, by model testing and theoretical calculations. Much attention has been given to the form, development and width of the surface subsidence trough that occurs during tunnel driving. The error curve profile developed as an empirical rule by Schmidt (1969) and followed by Peck (196 9).They also gave empirical relationships between the width of the trough and the dimensionles depth of the tunnel for broad types of intervening ground. XVIThese emprical rules provide useful practical guidelines, but the major problem is to understand how the surface settlements are related to the movements that are allowed to occur immediately around the tunnel and the properties of the varying types of ground lying between the tunnel and the ground surface. There remains a need for further information on ground movements about tunnels. As a result of the shotcrete technique in the last decade many observations of ground displacements between the tunnel and the ground surface have been made during tunnelling. These field data and ground movements made on tunnelling projects have been analysed by statistical methods to get a relationship to determine the settlements in terms of different parameters, These relationships have been restricted for certain soil types, depth and width of the tunnel. Finite element computer programs are being employed to study the soil behaviour around tunnels. It is a useful way of prediction of ground movements in lack of better methods. To get better results, an improved understanding of soil behaviour is necessary. For this, detailed laboratory tests are necessary to determine the soil parameters. Further work on the analysis and prediction of the subsurface deformation and displacements at the ground surface is in need of further development. Model tests provide helpful verification of theories. Recent important model tests are those performed by Lögters (1974) and by researchers in Cambridge, England (Atkinson, et al. (1974); Atkinson, Potts, Schofield (1977)). Improved soil models are needed to estimate the deforma tions around and above tunnels. More comprehensive studies are required in different types of ground. In addition to the studies thathave been made until today new experiments, more observations and investigations are needed to understand the behavior of the tunnels in loose-soft ground. In this thesis, investigations concerning the settlements and deformations at the ground surface and around the tunnel in relation with the settlement of the tunnel first in loose cohesionless soils were performed. XV HThe intention were to determine the effects of tunnel width, tunnel depth to width ratio and tunnel cross- section type to the settlements at the ground surface and around the tunnel. Tunnels with four different cross sections; circular, elliptical, rectangular and horseshoe - shaped were considerated. Test were carried out at five different tunnel depth to width ratios ; h/B=l,2, 3,4,5 and at three different tunnel widths; 10,5, 15, 19 cm. The model material, consisting of moist sand was placed and tested in a 51 x 104 x 122 cm iron box with plexiglas walls. To make the deformations more visible, black granulat was used at every 5 cm layer of model material. The settlement of the tunnel was performed by pulling out 1.2 cm thick boards. Total settlement of the tunnel was 6.0 cm. The settlements at the ground surface were measured with 6 computer-controlled extensometers. At the end of each test, deformations in the model body and at the surface were drawn on a transparent paper. In the first group, for each type of the given cross sections, tests were performed at h/B=l ratio, using one of the 10.5, 15 and 19 cm tunnel widths each time. In the second and third groups, tests performed in the first group were repeated for h/B=2 and h/B=3 ratios respectively. In the fourth group, at h/B=4 ratio, tests were performed for 10.5 and 15 cm tunnel widths. As the model box was not high enough, tests at 19 cm tunnel width could not be' performed. In the fifth group, at h/B= 5 ratio, tests were performed only for 10.5 cm tunnel width. As a total 4 8 tests were performed, Chapter 2 provides an overview and a brief summary of the studies performed about tunnels in soft soil. In addition a number of terms that are dealt with in more detail in the remaining chapters are introduced. XV 111Model tests, related model laws, model material and the methods used as well as the problems encountered in performing the tests are treated in chapter 3. The tests results are presented and discussed in chapter 4. General results are given in some detail in chapter 5. A Summary of the findings is presented below: - Maximum surface settlement increases linearly with increasing tunnel width at h/B=3 and 4 ratios. - At h/B « 2 ratios, maximum surface settlement increases with increasing h/B ratio whereas at h/B > 2 ratios, smax decreases with increasing h/B ratio. Tunnel depth is observed to have no effect on smax at h/B 5. 3 ratios. - It is pointed out that tunnel cross-section has no effect on smax for h/B s 2 ratios. In addition, it is observed that at h/B.*, 2 ratios, elliptical cross section gives smaller smax values. - sm /B increases linearly with increasing sc/B ratio at all cross-sections. At h/B=3,4 and 5 the relation between smax/B and sc/B ratios remains constant. - The effective width of the subsidence trough increases with increasing tunnel width. - In addition it must be emphassised that the shape of the subsidence trough changes with the depth of the tunnel. The width of the surface subsidence trough increases with increasing depth of the tunnel. The angle of repose is observed to change between 45-66.5°. The effective width of the tunnel is computed as 1.0-3.0 B for shallow tunnels and 2.0-5.5 B for deep tunnels. XIX