Ön konsolidasyon basıncının laboratuvar deneyleri ile belirlenmesi

Abstract

Suya doygun kohezyonlu zeminlerde, toplam oturmanın en önemli bileşeni, birincil konsolidasyon oturmasa, ol up« mühendislik tasarımında bu oturmanın değeri mümkün olduğunca gerçeğe yakın bir şekilde hesaplanmalıdır. Zeminin gerilme tarihçesinin konsol idasyon oturması üzerinde çok etkili olduğu da bilinmektedir. Zeminlerin konsolidasyon davranışını iyi anlayabilmek ve oturma hesaplarını doğru bir şekilde yapabilmek için, zeminin gerilme tarihçesinin bir göstergesi olan "ön konsolidasyon basıncı" nın belirlenmesi gerekir. Bu çalışmada, çamur konsol i dometre aleti ve Rowe Hücresi 'nde kontrollü şartlarda hazırlanan bilinen gerilme tarihçesine sahip numuneler üzerinde standart ödometre deneyleri yapılmış ve deney verileri değerlendirilerek beş değişik yöntemle ön konsolidasyon basıncı değerleri belirlenmiştir. Deney numunelerinin gerilme tarihçesi bilindiğinden, çeşitli yöntemlerle belirlenen Ön konsolidasyon basıncı değerleri ndek i hatalar tesbit edilmiş, yöntemler ayrıntılı bir şekilde tartışılmış ve ön konsolidasyon basıncının belirlenmesine yönelik bazı öneriler getirilmiştir.

In this thesis, laboratory tests have been conducted on cohesive soil samples prepared in slurry consol idometer and Rowe Consolidation Cell to determine the pre-consolidation pressure. Standard oedometer tests were carried out on soil samples with known stress history. Oedometer data then analysed with different methods reported in literature to determine the pre-consolidation pressure. Pre-consolidation pressures predicted with these well-known techniques then compared with the already known value of maximum past conso lidation pressure. In Chapter 1 of this thesis an introduction to the investigated problem is given. Chapter 2 deals with the consolidation concept and consolidation process. This is followed by methods suggested by several researchers for the determination of pre-consolidation pressure. Slurry consol idometer and Rowe Consolidation Cell are explained in detail, in Chapter 3. Sample preparation, loboratory tests and the evaluation of test results are also explained in this Chapter. An analysis of test data and conclusions of this study are outlined in Chapter 4. Primary consolidation settlement is the most significant component of the total settlement of a saturated cohesive soil. Therefore, primary consolidation should be precisely predicted in geotechnical desing. On the other hand, it is well known that stress history of a soil has great influence on the consolidation behaviour. In this context, stress history of a soil layer may be expressed in terms of over consolidation ratio or pre-consolidation pressure. viii In this study» laboratory samples with known stres: history have been prepared under controlled conditions and oedometer tests have been conducted on these samples. Soil samples were prepared in a slurry consolidometer and in Row© Consolidation Cell. A clay slurry having an initial water content of 1.5 to 2.5 times liquid limit were poured both in slurry consolidometer and Rowe Consolidation Cell and undisturbed soil samples were prepared by anisotropic consolidation of these slurry samples. Ultimate consolidation pressure thus the pre-consolidation pressure in the slurry consolidometer was selected as 100 kN/m. Corresponding ultimate consolidation pressures were 100 kN/m and 200 kN/m in Rowe Consolidation Cell. As can be seen from the above explanation three category tests have been carried out in this research work. Oedometer data obtained from standard oedometer tests have been evaluated by using five techniques in order to determine pre-consolidation pressure. These approaches are namely; Casagrande technique, Schmertmann technique, Janbu technique, Butterfield technique and Tavenas technique. First series of tests have been conducted on normally consolidated samples prepared in slurry consolidometer with a pre-consolidation pressure of 100 kN/m. Pre-consolidation pressure was then predicted with the implementation of five well-known methods Just mentioned above. Casagrande technique has given pre-consolidation values in the range of 95-125 kN/m for the samples consolidated with a maximum vertical effective stress of 100 kN/m. Schmertmann method makes the most accurate prediction with a maximum 20 % error. Tavenas method and Janbu method which considers (AH/H-c' ) V axes also make substantially good predictions. Janbu method adopting (M -©?' ) axes and Butterfield method have not provided reliable results for the first series of tests. Second series of tests have been conducted on normally consolidated samples prepared in Rowe Consolidation Cell with a maximum consolidation pressure of 100 kN/m. Five methods have also been tested to IX determine the already known value of pre-consolidation pressure. Casagrande method has given pre-consolidation pressures in the range of 90-170 kN/m which indicates a maximum 70 % difference from the known value. Corresponding differences are maximum 25 % in Janbu method which considers (AH/H-cr1 > axes and Tavenas method. Third series of tests have been carried on over consolidated samples that have OCR =2 and these samples were prepared in Rowe Consolidation Cell. Pre-consolidation pressures predicted with Casagrande method are in the range of 160-200 kN/m2 for this series of tests which gives indication to maximum 20 % difference from the known pre-consolidation pressure. Schmertmann method and Butterfield method have also given reasonable predictions. Janbu method which considersCAH/H-c' > axes and Tavenas method do not give V continuous curves. Therefore, determination of pre-consolidation pressure could not be possible with these methods. Conclusions of this thesis may be summarized as follows: 1) Clay samples having sufficient properties may be prepared with known stress history to be tested in laboratory research work by utilising the slurry consolidometer and Rowe Consolidation Cell. Preparation of clay samples in slurry consolidometer is quite simple and relatively similar to the stress history experienced by naturally deposited soils. On the other hand, Rowe Consolidation Cell has such advantages as drainage control, application of uniform and relatively high vertical stresses, pore water pressure measurement etc., although it doesn't fully reflect the natural sedimentation conditions. 2) Methods suggested by Casagrande and Schmertmann for the prediction of pre-consolidation pressure are valuable approaches. This study reveals that reasonable pre-consolidation values may be predicted with the implementation of these two methods. This is why the first method is widely accepted and used in geotechnical engineering at present. One of the shortcomings of Casagrande graphical construction is the determination of the minimum radius of curveture. Although not widely used Schmertmann method gives the most reliable pre-consol idation pressures due to the fact that this technique takes into account the construction of in-situ consolidation diagram. This research work has also revealed that the highest correlation was obtained for Schmertmann technique. This is well defined in Fig. 4. 2. Fig. 4. 2 shows that a correlation factor of R=0.82 is obtained for Schmertmann technique and the regression line is Just below the y=x üne which corresponds to 100 % correlation. 3) Stress -Deformation axes or Stress -Compressibility modulus axes may be used in Janbu method. Data analysed in stress-compressibility modulus axes provides the most sensible results for non-sensitive or little sensitive clays having high pre-consol idation pressure. It is recommended that oedometer test data should be analysed in stress-deformation axes for sensitive clays. Results of this research work reveal that stress-def ormation approach give reasonable pre-consol idation predictions in first and second series of tests where the pre-consol idation pressure is equal to 100 kN/m. Third category tests in which the pre-consol idation pressure is equal to 200 KN/m have given unreliable results. As to complete, the two techniques suggested by Janbu should be tested with further laboratory research. 4) Inconsistent soil behaviour has been observed for tests evaluated with Butterfield method although, Fig. 4- 3 shows a reasonably good correlation. Further tests should be carried out to test the Butterfield method. 5) For oedometer test samples that are not subjected to standard loading procedure (especially when the difference between any two loading level is high), Tavenas method doesn't usually give proper results. This situation is more important for test samples having a higher pre-consol idation pressure. As a result, very good predictions have been obtained for 100 kN/m xi vertical stress level by the application of Tavenas method. On the contrary, unreliable results have been obtained for 200 kN/m stress level. This valuable approach should also be studied with further oedometer tests having a special loading procedure.