Ayrık liflerle rasgele donatılı kum zeminlerin taşıma kapasitesinin CBR deneyleri ile araştırılması

Abstract

Donatılı zeminler, kullanımı hızla artan nispeten yeni bir geoteknik uygulamasıdır. Konvansiyonel zemin iyileştirme yöntemlerine oranla daha efektif, ekonomik ve hızlı çözümler sunmaktadır. Son yıllarda, donatılı zeminlerde sistematik donatı (geotekstil, geogrid v.b.) kullanımının yanı sıra mukavemet izotropisi sağlaması ve zeminle direkt karıştırılarak hazırlanması gibi bazı üstünlükleri sebebiyle ayrık liflerin rastgele donatı olarak kullanılması birçok araştırmaya konu olmuştur. Bu çalışmada, ayrık liflerle rastgele donatılı kum zeminlerin taşıma kapasitesi araştırılmıştır. Donatı tipi ve donatı oranı parametre alınarak bir dizi CBR (Kaliforniya Taşıma Oranı) deneyleri yapılmıştır. Ayrıca, ayrık liflerle rastgele donatılı zeminlere tatbik edilebilecek uygun sonlu elemanlar modeli araştırılmıştır. Bu deney ve analizlerden aşağıda belirtilen genel sonuçlar elde edilmiştir;. Zemin içerisine liflerin rastgele yerleştirilmesi, zeminin duktilitesini diğer bir ifadeyle, piston yükü-penetrasyon eğrisindeki pik gerilmelerive pik gerilmelerin mobilize olduğu penetrasyon miktarlarını artırmaktadır. Bu artış, kritik bir ağırlıkça donatı oranına kadar yaklaşık olarak sabit ve bu değerden itibaren üssel bir eğilim göstermektedir.. Zemin mukavemetindeki artış da benzer bir davranış sergileyerek kritik bir ağırlıkça donatı oranına kadar hemen hemen sabit bu değerden itibaren üssel bir eğilim göstermektedir.. Donatılı zeminlerde, pistonun belirli bir penetrasyon oranına kadar yük- penetrasyon eğrisinin eğiminde donatışız zeminlere oranla bir azalma olmaktadır.. Sonlu elamanlar analizlerinde rastgele donatı dağılımını sağlama açısında donatı konfigürasyonu önemli bir parametredir. Donatıların yatay yerleştirilmesi, başlangıç rijitliğinin artırılması açısından daha uygun sonuçlar vermektedir.

Earth reinforcement is an effective and reliable technique for increasing the strength and the stability of soils. Today, the technique is used with a variety of applications ranging from retaining structures and embankments for subgrade stabilization beneath footing and pavement. The material used for reinforcing purpose vary greatly in form (strips, sheets, grids, bars or fibers). The concept of reinforcing soil with discrete fibers extends to the early civilizations. As an accepted practice, sun-dried soil bricks which are reinforced with straw or other available fibers as a building material commonly used (Freitag, 1986). The use of fibers to reinforce soil as a primitive building material is still common but the mechanical behavior of randomly reinforced soils is not well-defined yet. In the last decade, however, an increasing number of study on the subject was conducted by several researchers, to investigate the engineering properties of randomly distributed fiber-reinforced soils. In comparison with oriented sheets of synthetic fabric, reinforcing soil with randomly distributed discrete fibers have some advantages. One of the main advantages is the maintenance of strength isotropy and the absence of potential planes of weakness that can develop parallel to oriented reinforcement. And the simplicity of the preparation method (simply added and mixed with the soil) could also be another main advantage. In the last decade many researchers attracted by these advantages. This study was undertaken to determinate the effects of reinforcement type and reinforcement weight ratio on the -randomly distributed fibers reinforced sand by performing CBR tests. Also, a proper model for finite element analysis was investigated. Chapter 1 introduces this study. The problem and the explanatory knowledge about the subject are stated. Chapter 2 presents a detailed literature review. To establish the background about the subject, the revolution of the randomly distributed fiber-reinforced sand was examined then, three different models, which were attempted to explain the mechanical behavior of the reinforced-sand, were stated. Force Equilibrium Model which is proposed by Gray and Ohashi (1983), accounts for the influence of much variables as the fiber modulus, diameter, initial orientation, and the elongation during shear; the skin friction between fiber and sand; and the angle of internal friction and relative density of the sand. The proposed model was viable only for oriented fibers and predetermined fiber area ratio which cross the failure surface. The calculation of the fiber area ratio could be possible for oriented fibers at direct shear tests, but the model needs modifications for randomly distributed fibers. Maher and Gray (1990), adopted the model on a statistical basis for estimation of fiber volume in a unit volume of sand. Then the fiber area ratio could be determined which is crossing any arbitrary surface from this fiber volume. Then the model managed to overcome the determination of fiber area ratio at failure surface. Another approach was Deformation Based Model, stated by Shewbridge and Sitar (1989; 1990; 1996), upon a series of laboratory tests. In the model, the strength increment depends on the shear zone pattern. The third approach was Energy Based Homogenization Model is stated by Michalowski and Zhao (1996). In the model the energy dissipation rate in the soil and fibers is calculated during the incipient deformation process, and it is equated to the work rate of the macroscopic (average) stress. In Chapter 4, the experimental study is presented. The material properties of the sand and the fiber reinforcement which are used in the tests are given. The test procedure and the test results are explained. The experimental study includes thirty CBR tests on randomly fiber-reinforced sand. Various types of fibers are employed as reinforcement. The weight proportion of the fiber and the fiber type are chosen as parameters. The effect of the parameters on;. stiffness,. ductility,. states of load and penetration values XI are investiagated. The soil, which the tests were performed on, is a poorly graded sand (SP) according to the Unified Soil Classification System (USCS). The soil chacteristics are summarized at Table 1. CAMEL YAF BMC 1-6 and CAMELYAF DEI -12 glass fibers and DUOMIX T20 fibrillated and DUOMIX M20 multi-filament polypropylene fibers were used as reinforcement (Table 2). The sand was placed in the standard CBR mold in five layers at a relative density of Dr = %60. The predetermined fiber amount was mixed by spatula until fibers were homogeneously distributed. The homogenity of the fiber distribution is inspected by visually. Table 1 The Test Sand Characteristics XII Table 2 The Fibers Characteristics The tests were conducted using a mechanically operated CBR apparatus. The penetration and the load values were recorded from dial gages which were positioned on the elevator table and the proving ring, respectively. The main findings obtained from experimental study are listed below:. In reinforced sands, the peak load values are obtained at higher penetration values. The relationship between load and penetration with the fiber weight fraction (p) is constant to a value of p approximately equal 0.2-0.75 % then increase exponentially. This fiber weight fraction was referred as Ireshold weight fraction. The rate of reinforced-unreinforced stress ratios (tfRpeak/aupeak) and weight fraction relation also has the similar behaviors (constant until to a value of p approximately equal 0.3-1.0 % then increase exponentially). The treshold weight fraction is depend on the fiber type.. Except Duomix F20, in all reinforced sand tests, the initial stiffness is found to be smaller than that in unreinforced sand. At higher penetration values (i.e. more than 10% of piston diameter), the composite system gets higher strength values than those of unreinforced sand.. The change of the reinforced-unreinforced stress ratios (üRpeak/crupeak) according to the fiber types, (p=l%; S/D=%40, S: Penetration value, D: Piston diameter) are listed below; Xlll - BMC 1-6 1.30 - DE1-12 1.70 - DuomixF20 7.20 - DuomixM20 3.10 (forp=0.50%) Chapter 4, introduces an attempt to establish a proper model for finite element analysis. In preliminary analyses, the effects of the compaction induced residual lateral earth pressure and the loading type (prescribed penetration or force loading) are investigated. In addition, an appropriate mesh design was tried to simulate the reinforcement configuration. It should be noted that modeling of states of stress and strain in reinforced sand during deformation and failure is complex and difficult. To establish more accurate models further studies are needed.