Yatay yükler etkisindeki büyük çaplı kazıkların davranışlarının belirlenmesi

Abstract

Kazıkların yatay yükler altındaki davranışlarının analizi için çeşitli yöntemler geliştirilmiştir. Yatak katsayısı metodu, kazıkların yatay yük taşıma kapasitelerinin belirlenmesinde çok yaygın olarak kullanılmaktadır. Zemine aktarılan basınçların elastik sınırlar içinde kaldığı durumlarda, lineer-elastik yatak katsayısı değerlerine ihtiyaç duyulur. Elastik sınırlar dışında, zeminin lineer olmayan davranışı ise, birbirinden bağımsız olan zemin yaylarının p-y eğrileri ile temsil edilmesi yoluyla modellenebilir. Yatak katsayısının elde edilebilmesi için, çeşitli araştırmacılar tarafından, birçok arazi ve laboratuvar teknikleri kullanılarak kolayca elde edilebilen zemin parametrelerine bağlı ampirik ve yarı ampirik bağıntılar önerilmiştir. Bu çalışmada, lineer-elastik yatay yatak katsayıları ile p-y eğrilerinin belirlenmesi amacı ile önerilen yöntemler anlatılmıştır. Yapılan araştırmalar, yatay yatak katsayısının, esas olarak, zeminin şekil değiştirme ve mukavemet özelliklerine bağlı olduğunu göstermekle birlikte kazık inşaa yöntemi, kazık yer değiştirmesinin büyüklüğü efektif düşey gerilme, zeminin rölatif sıkılığı ve kıvamı ile kazık boyutları ve eğilme rijitliği gibi çeşitli özelliklerden de etkilendiğini göstermiştir. Ayrıca, İstanbul Yeni Galata Köprüsü projesi kapsamında imal edilen 0 2000 mm çapında, 20 mm et kalınlığında ve maksimum 80 mt. boyundaki çelik boru kazıkların çakımı, köprü ekseni boyunca çıkarılan zemin profili ve kazıklar üzerinde yapılan yatay yükleme deneyleri anlatılmıştır. Yatay yatak katsayısı belirleme yöntemlerinden bazıları Karaköy ve Eminönü Bölgelerinde üzerinde yatay yükleme deneyleri yapılmış kazıklarda uygulanmış ve sayısal çözümleme sonuçları ile deney sonuçları karşılaştırılmıştır. Yapılan istatistiksel değerlendirme sonucunda, kohezyonlu zeminlerde drenajsız kayma mukavemeti, cu değerleri kohezyonsuz daneli zeminlerde SPT (N) darbe sayılarından elde edilen zemin elastisite modülü, Es değerleri ile her iki zemin türü içinde presyometre deneylerinden elde edilen Em Pl ve pf parametreleri kullanılarak yapılan çözümlemeler sonucunda belirlenen kazık başı fleksibilitelerinin deney sonuçlarıyla uyum içinde olduğu görülmüştür. Hesaplanan en büyük momentler ile bunların derinliklerinin ise kazık başı fleksibilitelerine kıyasla birbirlerine çok daha yakındır. Ayrıca, tersten hesap yapılarak yatay yatak katsayısının zemin elastisite modülü Es, drenajsız kayma mukavemeti Cu ve presyometre modülü Em değerlerine bağlı olarak elde edilmesinde kullanılmak üzere çeşitli bağıntılar önerilmiştir.

With urban growth and congested sites, the use of large diameter vertical piles to carry lateral loads has become more common. This type of piles are being widely used in offshore platforms, dolphins, jetties and at the foundations of the bridge abutments. There are a few general approaches that are available for the analysis of single piles subjected to lateral loads. In the Winkler or subgrade reaction approach of studying the lateral resistance of piles, the soil pressure (p) is related to the lateral deflection (y) through the modulus of subgrade reaction (kh ) ; P = kh. y The modulus of subgrade reaction has units of force/ ( length)2 If(kh)is multiplied by the length and diameter of a given pile segment, the equivalent spring stiffness is obtained. In this approach, the pile is considered to be supported by an array of uncoupled springs. These springs can be taken to be linear elastic, but more correctly they are taken to be nonlinear. The non-linear behaviour of the soil springs is represented by p-y curves which relate soil reaction and pile deflection at points along the pile length. Finite-element or finite-difference techniques can then be used to determine the response of the pile and spring system to applied loads. Using the beam on elastic foundation approach, basic equations have been developed for different variations of modulus of subgrade reaction (kh). The governing equation is EI (d4y / dz4 ) = - kh.y XXVIII where ; y = Pile deflection E = Young's modulus of pile, I = Moment of inertia of a pile cross-section z = Depth from surface Non dimensional coefficients have been given by several investigators for the solution of the laterally loaded pile problem to obtain deflection, moment, shear and so forth along the pile length. In problems dealing with lateral resistance of soil against piles, one needs to have a knowledge of the horizontal subgrade reaction modulus. There are several empirical and semiempirical relationships as well as charts and tables available for estimating the horizontal subgrade modulus. In addition, a variety of field and laboratory techniques have been used to determine the horizontal subgrade modulus, among which are standart penetration test, pressuremeter test, cone penetration test, plate load test, consolidation test, unconfined compression test and triaxial compression test. The most general form for either a horizontal or vertical modulus of subgrade reaction is, ks = As + Bs. zn where; As=constant for either horizontal or vertical members Bs=coef f icient for depth z =depth below ground n =exponent to give (ks ) the best fit The relationship between pressure and deflection at any point along a pile is nonlinear. A number of approaches have been developed to account for nonlinear behaviour. The most general approach to account for nonlinear soil behaviour is the p-y approach developed by Reese and his coolleagues. Most of the existing methods for obtaining p-y curves are highly empirical. The p-y curves are dependent on pile geometry, soil properties and depth below ground surface. The p-y curves can be developed from results of pressuremeter tests, undrained strength data and results of the field load tests. The pressuremeter offers an almost ideal in-situ modelling tool for determining directly the p-y curves for a pile. As the pressuremeter can either be driven or self-bored in to the soil, the results can be used to model either a displacement or a non-displacement pile. XXIX Even though the value assigned to the horizontal subgrade modulus plays a significant role in the computations of soil resistance of laterally loaded piles, a rather wide range of values is available in literature depending upon the equation, chart or table used to obtain the modulus. The accuracy of elaborate methods of analysis for lateral load carrying capacity is often controlled by the accuracy of the modulus value used in the computations. In selecting a procedure to obtain the modulus for lateral pile capacity computations, the engineer must strike a balance between accuracy and simplicity. One of the simplest methods of computing the modulus is by means of empirical and semiempirical relationships, which relate the modulus to other known or easily obtainable soil properties. In this study, the available empirical and semiempirical relationships as well as tables and charts for predicting the coefficient of horizontal subgrade reaction are reviewed. The coefficient of horizontal subgrade reaction is influenced by many factors among which are the magnitude of pile deflection, the effective overburden pressure, the relative density of soil, ground water condition, nature of applied load (static and dynamic) and the properties and shape of the pile section. In the fourth chapter of this study, information about the piles, soil conditions and the lateral load tests of the New Galata Bridge are explained. The piles of the New Galata Bridge were spirally welded steel pipe piles of St. 52. 3. The pile diameter was 2000 mm and the wall thichness was 20 mm. Due to the soil conditions some of the piles were driven while others were driven and bored. Due to the absence of a gravel layer on Karaköy side, the abutment and the piles on the first two axes on this side were driven open ended to the rock layer. The rock layer was bored to minimum 1,5 times of the pile diameter and a bottom plug was formed. All other piles were closed ended and driven to the refusal in the gravel layer. In the area of the bridge crossing the shores are completely covered by thick layers of man-made fill. The thickness of the fill is more than 40 m. along the Eminönü shore, and more than 30 m. along the Karaköy shore. The thickness of the fill decreases with increasing distance from the shoreline. The borings which were carried out along the bridge axis indicated that the man-made fill is underlain by sedimentary alluvium layer consisting of sand, silt and clay in varying proportions. In general the alluvium is described as graydark gray fat organic silty clay of marine origin. It is maximum thickness along the bridge axis is about 35 m. At Eminönü XXX side, the alluvium is underlain by a few meters of weathered shale. Towards the Karaköy side the alluvium is underlain by cobbly gravel. The sandstone bedrock underlies the whole area of the bridge crossing. Four lateral load tests were performed by means of hydraulic jacking system apart pairs of piles at specified locations. The distance between the axis of the pairs of piles were 10 m. in sea section and 6.30 m. in abutments. Increments of loading were applied in two cycles as required by the specification. The load applied was controlled by two calibrated manometers. Lateral pile head movements were measured using independent beams with pointers to scales attached laterally to each pile head. At the end of this part, the results of lateral load tests according to the given loading program are summarized. In the fifth chapter, linear-elastic horizontal subgrade reaction variations along the laterally loaded test piles lenght at the Karaköy Approach Spans (Piles No: 7'.1 and 7 '. 2 ) ; Eminönü Approach Spans (Piles No : 4. 3 and 4.4); Karaköy Abutment (Piles No: 0'.8 and 0'.9) and Eminönü Abutment (Piles No: 0.8 and 0.9) are calculated from empirical and semi-empirical equations which are proposed from elastic modulus of the soil, Es by Glick (1948), Terzaghi (1955) Vesic (1961), Francis (1964), Poulos (1971), DIN 4014 (1977) Pyke & Beikae (1984) and Bowles (1989) ; from undrained shear strength of the soil, cu by Davisson (1970) and Robinson (1979) and from pressuremeter test parameters Em, pl, pf by Menard (1962), Chen (1978), Baguelin et al (1978) and Briaud et al (1984). The variation of the horizontal subgrade reaction with depth is taken linear and standart penetration test (SPT), field veyn test, pressuremeter test (PMT) odometer test and direct simple shear test results are used in calculations. Non-linear p-y curves based on field pressuremeter tests are calculated by using semi-empirical equations proposed by Baguelin(1978). In the sixth chapter, the pile head deflections which are measured by lateral load test at Karaköy Approach Span Karaköy Abutment, Eminönü Approach Span and Eminönü Abutment are compared with the values computed using various linear- elastic horizontal subgrade reaction variations along the test pile length. SAP 90 computer program is used to predict behaviour of the pile under the test conditions. Futhermore non-linear behaviour of the pile-soil system is solved by LUSAS computer program and the results are summarized. In the seventh chapter, the results obtained from the various horizontal subgrade reaction variations with dept are evaluated by using statistical analysis STATGRAPHICS computer program is used for this aim. XSSI Finally, the conclusions deduced are summarized in chapter eight. It is found that the pressuremeter test is very suitable for use in calculating horizontal subgrade reaction variations with depth and establishing p-y curves for laterally loaded piles. In cohesive soils, undrained shear strength parameters, cu which have been obtained, from field and laboratory tests and in granular soils, standart penetration tests values give better results and computed pile head deflections are in a harmony with the test results. Maximum moments and their depths are closer to each other than the pile deflections.