Göçme durumuna ulaşmamış kazık yükleme deneylerinde nihai kazık taşıma kapasitesinin belirlenmesi
Göçme durumuna ulaşmamış kazık yükleme deneylerinde nihai kazık taşıma kapasitesinin belirlenmesi
Dosyalar
Tarih
2020-06-15
Yazarlar
Dalkılıç, Berkay
Süreli Yayın başlığı
Süreli Yayın ISSN
Cilt Başlığı
Yayınevi
Fen Bilimleri Enstitüsü
Institute of Science and Technology
Institute of Science and Technology
Özet
Kazık yükleme deneyleri, nihai kazık taşıma kapasitesinin belirlenmesi amacıyla işyeri sahasında bire bir ölçekte uygulanan güvenilir deneylerdir. Çeşitli araştırmacılar tarafından önerilen yöntemler doğrultusunda teorik olarak kazık taşıma kapasitelerinin hesaplanması mümkün olmaktadır. Ancak, temel altı kazıkların tasarımında yapılan öngörüler ve belirlenen parametreler bazı belirsizlikler içermektedir. Zeminin homojen olmayan yapısı, zemin araştırma çalışmaları esnasında gerçekleşebilecek hatalar sonucu araziden alınan numunelerin sahadaki gerçek durumu temsil etmemesi, tasarımcı tarafından yapılan kabullerin kişiden kişiye farklılık göstermesi gibi nedenlerden dolayı teorik yöntemler ile ulaşılan kazık taşıma kapasitelerinin doğrulanması gerekmektedir. Bu sebeple; tasarımdaki olası hataların düzeltilmesi, aşırı güvenli dizaynların önüne geçilerek daha ekonomik çözümler üretilmesi amacıyla kazık yükleme deneyleri ile gerçek kazık taşıma kapasitelerinin elde edilmesi büyük önem arz etmektedir. Kazık yükleme deneylerinde, test kazığının göçme durumuna ulaşıncaya kadar yüklenmeye devam edilmesi halinde kazık taşıma kapasitesi yalnızca yük-oturma grafiği yorumlanarak bile kolaylıkla belirlenebilmektedir. Ancak; bazı deneylerde, kazığın göçmesine neden olacak büyüklükteki yükü, test kazığına aktarabilmek için gereken reaksiyon sisteminin (çekme kazıkları, test kirişleri, hidrolik krikolar vb.) kurulumu pratik olarak mümkün olmamaktadır. Bu doğrultuda; tezin konusu olan göçme durumuna ulaşmamış kazık yükleme deneylerinde nihai kazık taşıma kapasitesinin doğru şekilde belirlenmesi gerekmektedir. Aksi halde, yapımı maliyetli ve zaman alıcı bu deneyler esas amacına ulaşmamış olacaktır. Literatür çalışması kısmında tek bir kazık taşıma kapasitesi değerine ulaşılabilmesi için 40'tan fazla farklı yöntem önerildiği anlaşılmaktadır. Bu doğrultuda; yöntemlerin olumlu ve olumsuz yönlerinin değerlendirilerek hangi yöntemin tercih edileceği önem kazanmaktadır. Tez çalışmasında; büyük bir kısmı göçme durumuna ulaşmamış 25 farklı kazık yükleme deneyi için, farklı araştırmacılar tarafından önerilen ve dünyada genel kabul görmüş yöntemler kullanılarak kazık taşıma kapasiteleri hesaplanmıştır. Ayrıca; göçme durumunun deney sonuçlarının yorumlanması üzerindeki etkisine de değinilmesi amacıyla, göçme yüküne kadar devam ettirilmiş bazı deneylere de yer verilmiştir. Deneylerin değerlendirilmesinde; Brinch Hansen %80, Brinch Hansen %90, Butler - Hoy, Cambefort - Chadeisson, Chin - Kondner, Corps of Engineers, Davisson, De Beer, Decourt, Fuller - Hoy, Hirany - Kulhawy, Mazurkiewicz, Teğet (Mansur - Kaufman) ve Van Der Veen olmak üzere 14 farklı yöntem kullanılmıştır. Ulaşılan veriler ışığında; göçme durumuna kadar devam ettirilmemiş kazık yükleme deneylerinin değerlendirilmesinde Fuller - Hoy, Hirany - Kulhawy, Brinch Hansen %90 ve Davisson yöntemlerinin uygulanabilirliklerinin çok düşük olduğu, tez çalışmasında açıklanan diğer yöntemlerin ise yüksek oranda başarılı sonuçlar verdiği görülmüştür. Tüm deney sonuçları göz önüne alındığında; en yüksek taşıma kapasitesi değerine Decourt yöntemi ile, en düşük taşıma kapasitesine ise De Beer yöntemi ile ulaşıldığı anlaşılmıştır. Bunun dışında; deneyler ve farklı yöntemler için çeşitli veri havuzları oluşturularak istatistiksel analizler yapılmış, standart sapmalar üzerinden varyasyon katsayıları hesaplanmıştır. Bu varyasyon katsayıları doğrultusunda belirsizlikler yorumlanmıştır. Buna göre farklı yöntemlerin; göçme durumuna kadar yüklemeye devam edilen deneylerde, göçme durumuna ulaşmamış deneylere kıyasla birbirine çok daha yakın sonuçlar verdiği belirlenmiştir. Yöntemlerin belirsizlikleri incelendiğinde ise genel olarak kendi arasında en tutarlı sonuçların Butler - Hoy, Mazurkiewicz ve Corps of Engineers yöntemleri ile, en kararsız sonuçların ise Brinch Hansen %80, Chin - Kondner ve Decourt yöntemleri ile elde edildiği gösterilmiştir. Ayrıca, aynı deney verileri üzerinde farklı yöntemler uygulanarak elde edilen kazık taşıma kapasiteleri arasında %300' varan büyük mertebe farkları olduğu görülmüştür. Bu doğrultuda, yöntemler karşılaştırılarak her yöntem için ayrı güvenlik katsayıları önerilmiştir. En büyük güvenlik katsayısı değerini Decourt ve Chin - Kondner yöntemleri, en küçük güvenlik katsayısı değerini ise De - Beer, Cambefoty - Chadeison ve Teğet (Mansur - Kaufman) yöntemleri almıştır.
Pile loading tests are reliable experiments which are performed on one to one scale in the workplace that represents the real soil/rock structure to determine the ultimate pile bearing capacity and confirm the assumptions and calculations made on the design state. In line with the methods which are proposed by various researchers, it can be possible to calculate the pile bearing capacities theoretically. However, the predictions and parameters which are determined in the design state of pile foundations contain some uncertainties. Pile bearing capacities which are reached by theoretical methods should be verified due to the inhomogeneous structure of the ground, the samples taken from the field that may not represent the real state as a result of errors that may occur during the soil survey studies, and the assumptions made by the designer may differ from person to person, even with the same datas, the different pile bearing capacities can be reached. For these reasons, it becomes very important to obtain real pile bearing capacities via pile loading tests in order to correct possible errors and to produce more economical solutions by preventing over designs. In the pile loading tests, the pile bearing capacity can be easily determined even by just interpreting the load - settlement graph only in case of being continued loading of the test pile until it reaches the failure state. However, generally in loading tests on large diameter bored piles, it is practically not possible to install the system (tensile piles, reaction beams, hydraulic jacks etc.) which is required to transfer the load that is enough to cause the pile fail into the test pile. Accordingly, in pile loading tests, being the subject of the thesis, which have not reached the failure state, the ultimate pile bearing capacity must be determined correctly. Otherwise, these experiments which are costly and time consuming will not reach their main purpose. In the literature research section, it can be seen that there are more than 40 different methods proposed by different researchers to obtain one single ultimate pile bearing capacity. But most of them are inadequate in evaluating the results because of the settlement criterions they suggest. So, it becomes very important to decide which method to use by judging their positive and negative aspects. In this thesis study, pile bearing capacities have been calculated for the twenty-five different pile loading tests via using methods which are recommended by different researchers and generally accepted and used in the world. The majority of loading tests (20 of the 25 tests) are choosed from that have not reached the failure state. Also, five tests which have been continued until the failure load have been included for the purpose of mentioning the impact of the failure state on interpretation of the experiment results. In the evaluation of the experiments, fourteen different methods have been used, namely Brinch Hansen %80, Brinch Hansen %90, Butler - Hoy, Cambefort - Chadeisson, Chin - Kondner, Corps of Engineers, Davisson, De Beer, Decourt, Fuller - Hoy, Hirany - Kulhawy, Mazurkiewicz, Tangent (Mansur - Kaufman) and Van Der Veen. In the light of the datas obtained, it has been seen that the applicability of the Fuller - Hoy, Hirany - Kulhawy, Brinch Hansen %90 and Davisson methods have been very low in the evaluation of the pile loading tests which have not been continued to the failure state. The main reason of that is the settlement conditions described by these methods cannot be reached for the tests that have not reached the failure state. The other methods which have been explained in this thesis have yielded highly successful results. Considering all test results; it has been understood that the highest bearing capacity values have been reached by Decourt and Chin - Kondner methods. In these two methods, the pile bearing capacities are calculated by matematical extrapolations and the results are very similar. The lowest bearing capacities have been reached by De Beer method. Apart from these, statistical analyses have been done by creating various data pools for tests and different methods and variation coefficients have been calculated via standard deviations. Uncertainties have been interpreted in line with these variation coefficients. Accordingly, it has been determined that different methods have given much closer results in tests having been continued until the failure state compared to experiments which have not reached the failure state. According to this, it is recommended that the loading tests shouldn't be ended in very small settlements. The loading tests should be continued as much as the reaction system allows. By this way, as the test load approaches to the failure load, more consistent results can be obtained and the ultimate pile bearing capacities are going to interpreted more accurately. When the uncertainties of the methods are examined, it has been shown that the most consistent results among themselves are obtained through Butler - Hoy, Mazurkiewicz and Corps of Engineers methods. The most unstable results are obtained through Brinch Hansen %80, Chin - Kondner and Decourt methods on the pile loading tests which have not been continued to the failure state. In addition, it has been observed that there are big differences, approximately %300, in pile bearing capacities which have been obtained through different methods for the same pile loading test datas. So, it is clear that the same values of safety coefficients cannot be used for all methods. For this reason, by comparing the methods and using Fellenius' suggestions, separate factors of safety have been interpreted for each method. The maximum factor of safety is proposed for the Decourt and Chin - Kondner methods (3.00) which gives the highest bearing capacities, and the minimum factor of safety for the De - Beer, Cambefort - Chadeisson and Tangent (Mansur - Kaufman) methods (1.60) which gives the lowest bearing capacities. These safety coefficient are suggested only for the loading tests that have not reached the failure state because of it has been obsereved that in the experiments that have reached the failure state, different methods already gives similar results. As a result, the methods can be grouped in three different approaches: extrapolation with matematical relations, graphical applications and determining a limit settlement value. The assumptions of the designer becomes very important especially for the graphical methods. For example, how to divide the y-axis (settlement axis) for Mazurkiewicz method or the tangent drawn on uncomplete plastic part of the load - settlement graph for Tangent (Mansur - Kaufman) method have a big impact on determining the pile bearing capacity. When determining the ultimate pile bearing capacity, it is suggested that it will be a more realistic approach to use more than one method and applying the recommended safety factors for each method. After appliyng the recommended safety factors, the average of the results can be taken as an ultimate pile bearing capacity. Also, it should be noted that the settlements corresponding to the ultimate pile bearing capacity value is limited by the amount of settlement value allowed by the structure that is carried by the deep foundation.
Pile loading tests are reliable experiments which are performed on one to one scale in the workplace that represents the real soil/rock structure to determine the ultimate pile bearing capacity and confirm the assumptions and calculations made on the design state. In line with the methods which are proposed by various researchers, it can be possible to calculate the pile bearing capacities theoretically. However, the predictions and parameters which are determined in the design state of pile foundations contain some uncertainties. Pile bearing capacities which are reached by theoretical methods should be verified due to the inhomogeneous structure of the ground, the samples taken from the field that may not represent the real state as a result of errors that may occur during the soil survey studies, and the assumptions made by the designer may differ from person to person, even with the same datas, the different pile bearing capacities can be reached. For these reasons, it becomes very important to obtain real pile bearing capacities via pile loading tests in order to correct possible errors and to produce more economical solutions by preventing over designs. In the pile loading tests, the pile bearing capacity can be easily determined even by just interpreting the load - settlement graph only in case of being continued loading of the test pile until it reaches the failure state. However, generally in loading tests on large diameter bored piles, it is practically not possible to install the system (tensile piles, reaction beams, hydraulic jacks etc.) which is required to transfer the load that is enough to cause the pile fail into the test pile. Accordingly, in pile loading tests, being the subject of the thesis, which have not reached the failure state, the ultimate pile bearing capacity must be determined correctly. Otherwise, these experiments which are costly and time consuming will not reach their main purpose. In the literature research section, it can be seen that there are more than 40 different methods proposed by different researchers to obtain one single ultimate pile bearing capacity. But most of them are inadequate in evaluating the results because of the settlement criterions they suggest. So, it becomes very important to decide which method to use by judging their positive and negative aspects. In this thesis study, pile bearing capacities have been calculated for the twenty-five different pile loading tests via using methods which are recommended by different researchers and generally accepted and used in the world. The majority of loading tests (20 of the 25 tests) are choosed from that have not reached the failure state. Also, five tests which have been continued until the failure load have been included for the purpose of mentioning the impact of the failure state on interpretation of the experiment results. In the evaluation of the experiments, fourteen different methods have been used, namely Brinch Hansen %80, Brinch Hansen %90, Butler - Hoy, Cambefort - Chadeisson, Chin - Kondner, Corps of Engineers, Davisson, De Beer, Decourt, Fuller - Hoy, Hirany - Kulhawy, Mazurkiewicz, Tangent (Mansur - Kaufman) and Van Der Veen. In the light of the datas obtained, it has been seen that the applicability of the Fuller - Hoy, Hirany - Kulhawy, Brinch Hansen %90 and Davisson methods have been very low in the evaluation of the pile loading tests which have not been continued to the failure state. The main reason of that is the settlement conditions described by these methods cannot be reached for the tests that have not reached the failure state. The other methods which have been explained in this thesis have yielded highly successful results. Considering all test results; it has been understood that the highest bearing capacity values have been reached by Decourt and Chin - Kondner methods. In these two methods, the pile bearing capacities are calculated by matematical extrapolations and the results are very similar. The lowest bearing capacities have been reached by De Beer method. Apart from these, statistical analyses have been done by creating various data pools for tests and different methods and variation coefficients have been calculated via standard deviations. Uncertainties have been interpreted in line with these variation coefficients. Accordingly, it has been determined that different methods have given much closer results in tests having been continued until the failure state compared to experiments which have not reached the failure state. According to this, it is recommended that the loading tests shouldn't be ended in very small settlements. The loading tests should be continued as much as the reaction system allows. By this way, as the test load approaches to the failure load, more consistent results can be obtained and the ultimate pile bearing capacities are going to interpreted more accurately. When the uncertainties of the methods are examined, it has been shown that the most consistent results among themselves are obtained through Butler - Hoy, Mazurkiewicz and Corps of Engineers methods. The most unstable results are obtained through Brinch Hansen %80, Chin - Kondner and Decourt methods on the pile loading tests which have not been continued to the failure state. In addition, it has been observed that there are big differences, approximately %300, in pile bearing capacities which have been obtained through different methods for the same pile loading test datas. So, it is clear that the same values of safety coefficients cannot be used for all methods. For this reason, by comparing the methods and using Fellenius' suggestions, separate factors of safety have been interpreted for each method. The maximum factor of safety is proposed for the Decourt and Chin - Kondner methods (3.00) which gives the highest bearing capacities, and the minimum factor of safety for the De - Beer, Cambefort - Chadeisson and Tangent (Mansur - Kaufman) methods (1.60) which gives the lowest bearing capacities. These safety coefficient are suggested only for the loading tests that have not reached the failure state because of it has been obsereved that in the experiments that have reached the failure state, different methods already gives similar results. As a result, the methods can be grouped in three different approaches: extrapolation with matematical relations, graphical applications and determining a limit settlement value. The assumptions of the designer becomes very important especially for the graphical methods. For example, how to divide the y-axis (settlement axis) for Mazurkiewicz method or the tangent drawn on uncomplete plastic part of the load - settlement graph for Tangent (Mansur - Kaufman) method have a big impact on determining the pile bearing capacity. When determining the ultimate pile bearing capacity, it is suggested that it will be a more realistic approach to use more than one method and applying the recommended safety factors for each method. After appliyng the recommended safety factors, the average of the results can be taken as an ultimate pile bearing capacity. Also, it should be noted that the settlements corresponding to the ultimate pile bearing capacity value is limited by the amount of settlement value allowed by the structure that is carried by the deep foundation.
Açıklama
Anahtar kelimeler
Fore kazık,
Bored pile