Akademik Veri Yönetim Sistemi
Tezin Türü: Doktora
Tezin Yürütüldüğü Kurum: İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, inşaat mühendisliği, Türkiye
Tezin Onay Tarihi: 2018
Tezin Dili: İngilizce
Öğrenci: İNCİ SÜT ÜNVER
Danışman: Erdal Çokça, Musaffa Ayşen Lav
Özet:
STATIC AND CYCLIC PROPERTIES OF EXPANSIVE CLAYS TREATED WITH LIME AND FLY ASH WITH SPECIAL REFERENCE TO SWELLING AND RESILIENT MODULI
SUMMARY
Water intake capacity of the clay type of soils generally increases with increasing plasticity index. Soils with high plasticity and high swell potential have great volume changes due to water content fluctuations in an unstable manner. As a result of this expansion, the pavements and/or lightweight structures such as transportation facilities, which are settled on that kind of subsoils, are damaged. Expansive soils can be improved by adding chemical stabilizers such as lime, fly ash, micro cement, etc. Chemical stabilizers improve swelling and swell pressure characteristics of expansive clays as a result of cation exchange and flocculation-agglomeration. Chemical stabilizers also help to improve bearing capacity of such soils.
The resilient modulus value, on the other hand, provides a basic relation between the stress and strain of the material in order to perform the structural analysis of the layered pavement systems under cyclic conditions. This value enables to define highway materials under various conditions such as water content, density etc. and under different stress conditions that exemplify live wheel loads.
Lime and fly ash are the most popular agents used in chemical stabilization method. However, it is not very clear whether these agents are effective enough or not effective at all in very highly or extremely highly plastic clays. Moreover, swell and shrinkage characteristics of the lime and fly ash stabilized soils are not well-known when compared to the strength behavior of the treated soils. Another not well-known aspect of the subject is cyclic properties of the native expansive soils and stabilized soil-agent mixtures. It is believed that the resilient modulus properties are needed to be investigated in this context. Finally, the effect of mineralogical composition of the clay soils and chemical components of the agents are believed to be investigated as far as the swelling is concerned. In this context, both treated soils and the lime and fly ash agents used in this research are identified by means of Rietveld, XRD and XRF tests.
In this research, static and cyclic properties of the two natural clays, which are extremely highly and very highly plastic expansive soils, are investigated by utilizing lime and fly ash agents for stabilization. Static research has been executed with special reference to swelling behavior. Resilient modulus properties have been examined under variable conditions within the scope of dynamic research.
Two different soil samples, which are taken from different sites of Ankara, have been investigated. They are namely, Esenboğa and Bilkent Clays. Esenboğa Clay is an extremely highly plastic clay (PI=101%). Esenboğa Clay samples have been taken from clay deposits in the Akyurt district of Ankara, near Esenboğa Airport. Bilkent Clay is a very highly plastic clay (PI=76%). Bilkent Clay samples have been taken from clay deposits of the Bilkent Atatürk Hospital district of Ankara, near Ankara-Eskişehir State Highway. Both disturbed and undisturbed soil samples have been taken from both sites in order to perform laboratory tests and hydrated lime and fly ash (Class C) have been used as chemical stabilizers.
The soil mineralogy of the clays used in this research has been determined by XRD and Rietveld tests that are supported by the petrographical, XRF and SEM analyses. Detailed mineralogical identifications in accordance with the ASTM and quantitative mineral analysis by utilizing appropriate softwares and databases are conducted by means of the standard (2°-70°), normal (2°-30°), ethylene glycol, 300°C thermal and 550°C thermal process imaging. For the chemical composition determination of the clay samples and the agents that are used in this study (i.e. lime and fly ash), XRF tests have been conducted. Samples are prepared and analyzed as pellet structures after being dried at 65°C during one night and then at 105°C during 2 hours in XRF tests. These samples are studied within the range of Bor-Uranium and the results are presented in the form of oxide.
Firstly, reference tests have been carried out on untreated natural soil samples. Then, both clays are mixed with the lime agent at different percentages of dry weight of the soil (i.e. 1%, 3%, 5%, 7% and 9%). After that, the clays are mixed with the fly ash agent at different percentages of dry weight of the soil (i.e. 5%, 10%, 15%, 20% and 25%).
Methylene blue tests have been performed on untreated and lime-treated Esenboğa and Bilkent Clays for each percentage in order to determine the activity of the clay and also to have an idea about the swell potential, index properties and specific surface area.
Index tests have been conducted on untreated and treated samples. The oedometer tests have been performed to determine the swelling and compressibility characteristics of the soil samples. The unconfined compression tests have been executed for both non-soaked and soaked conditions. These statical tests have also been performed for Esenboğa Clay with lime agent for each percentages prepared for 7, 28 and 56-days curing periods. All statical tests of Bilkent Clay have been conducted with samples which are not cured.
Moreover, resilient modulus tests, which are essentially a cyclic version of a triaxial compression test, are performed on lime-treated samples for different curing periods to better simulate the actual traffic loading more accurately. Cyclic load application in the test here is thought to represent the traffic loading more realistically.
The Esenboğa and Bilkent Clays, which are mixed with the lime agent at different percentages of dry weight of the soil (1%, 3%, 5%, 7% and 9%), have been tested for the determination of resilient modulus values. Here, resilient moduli have been found for soil samples prepared for 7, 28, 56 and 90-days curing periods in each lime percentages.
The microstructural development of lime-treated clay sample (only 7%) taken from resilient modulus test sample and then waxed, has been imaged by SEM analyes with a curing period of 30 months for Esenboğa Clay. SEM analyses have been performed for natural Esenboğa Clay and Bilkent Clay as well.
Engineering properties of the improved soil samples have been determined. Index, strength and swelling properties of the soil samples have been found. All of the results are evaluated and the changes in all the above mentioned properties especially with regard to swell potential and compressive strength and the time effect on improvement
xxxiv
and mean resilient modulus values of the improved soil samples have been interpreted in this thesis. Optimum values have also been evaluated for swell potential and compressive strength. The correlations between the agent percentages and index properties, the swell percentage, the swell pressure, time effect, compressive strength and resilient modulus are presented.
As a result of mineralogic analyses of the soil samples, it is understood that the Esenboğa Clay is composed of mainly illite (about 87%). However, it is determined that the amount of smectite minerals are not negligible (about 4%). It is interesting to note, on the other hand, that the Bilkent Clay is completely composed of palygorskite (attapulgite) mineral.
The optimum lime content is about 3% as far as the swelling is concerned in Esenboğa Clay. In a design of sub-base of surficial structures (i.e. highways, railways, etc.), this optimum lime content can be taken as a reference. Swell tests should be executed with this reference value. In case of not satisfying the design criteria, this optimum lime content should be increased. For instance, laboratory swell percentage tests show that about 8% lime content for Esenboğa Clay may just be sufficient for design criteria according to Highway Standards in Türkiye. Laboratory tests show, on the other hand, that the fly ash agent may not be sufficient alone for swelling design criteria in such type of extremely and/or very highly plastic clays. Lime agent effect on swelling behavior, therefore, is much more effective than the effect of fly ash agent. For instance, swell pressures are about 6.5 times greater in fly ash improvement with 20 to 25% amounts than the pressures found in lime improvement with 7 to 9% amounts in Esenboğa Clay.
When the variation of unconfined compressive strength with lime content for non-soaked and soaked conditions is considered without curing, it is noticed that the unconfined compressive strengths are greater in non-soaked condition up to 5% lime content. Starting from 7% however, the strengths are greater in soaked condition due to self-hardening mechanism which can be called also as pozzolanic reactions.
The unconfined compressive strengths are always greater in non-soaked condition for fly ash treatment without curing. It is seen that 15% fly ash content is optimum since the unconfined compressive strengths decrease at 20% and 25% fly ash contents for both non-soaked and soaked samples.
Time effect of lime stabilization in the extremely highly plastic Esenboğa Clay has been investigated in this research. The soil samples for cure periods of 7, 28 and 56 days have been tested as mentioned above.
For the curing periods of 28 and 56 days, 5% lime content is the optimum content at which the swell percentage is decreased to null in Esenboğa Clay. It is interesting to note that 28 days curing period, which resembles the optimum time for strength gain in concrete, is quite adequate to get allowable swell percentages at about 5% lime content. Then, it can be concluded that 56 days curing period may not be necessary as far as the swell percentage criterion is concerned. This result is also valid in swell pressure tests so that 5% lime content is sufficient to decrease the swell pressures considerably for all curing periods.
Soaked and non-soaked samples have been tested to get unconfined compressive strengths in Esenboğa Clay with different curing periods. It has been concluded that the unconfined compressive strengths increase dramatically as compared to non-cured strengths for all cure periods especially for non-soaked samples. 28-day curing period can be accepted as optimum. Unconfined compressive strengths of soaked samples are always lower than the ones in non-soaked condition in all lime contents for curing periods of 7, 28 and 56 days.
The optimum lime and fly ash contents for Bilkent Clay are about 3% and 10% respectively as far as the swelling is concerned. Since lime is more effective than the fly ash, it is highly probable that the design criteria are to be satisfied at reasonable lime contents. For instance, laboratory swell percentage tests show that 9% to 10% lime content for Bilkent Clay may be sufficient in a transportation sub-structure design as far as the swelling criteria in Highway Standards in Türkiye are concerned. Laboratory tests show, on the other hand, that the fly ash agent may not be sufficient alone for swelling design criteria in such type of very highly plastic clays.
The effectiveness of lime agent on swell behavior can also be observed in swell pressure tests in Bilkent Clay. The effect of lime on swell pressure is maximum in lime content increase from 1% to 3%. This is compatible with the result of swell percentage tests. Swell pressures, on the other hand, are about 2.7 times greater in fly ash improvement with 20 to 25% amounts than the pressures found in lime improvement with 7% to 9% amounts in Bilkent Clay.
When the variation of unconfined compressive strength with increasing lime content for non-soaked and soaked conditions is considered without curing, it is noticed that the unconfined compressive strengths are greater in non-soaked condition up to 7% lime content. Starting from 7%, the strengths are greater in soaked condition due to pozzolanic reactions.
The unconfined compressive strengths are always greater in non-soaked condition for fly ash treatment in Bilkent Clay.
The methylene blue tests have been performed on lime-treated samples for both clays. The results obtained are quite compatible with the results mentioned above. That is to say, the swell pressure and the swell percentages decrease with the increase in lime content, the methylene blue values (MBV) and activity decrease as well.
When the variation in the mean resilient modulus values are considered for the extremely highly plastic Esenboğa Clay (PI=101%), 3% lime content seems quite effective for both non-cured and cured samples. 56-days curing period gives maximum mean resilient modulus values in this clay for which the 90-days can be accepted ineffective.
In Bilkent Clay which is very highly plastic (PI=76%), 3% lime content can be accepted effective at early stages (i.e. at non cured and 7 days cured conditions) as far as the mean resilient modulus values are concerned. However, 7% lime content gives generally maximum increases in resilient modulus values at later cure periods of 28, 56 and 90 days. 90-days curing period gives always maximum resilient modulus value in the tests for Bilkent Clay. It should be noticed here that the Bilkent Clay is composed of 100% palygorskite clay mineral which may influence the results.
The results achieved in this research show that lime agent alone can be utilized quite effectively in the extremely highly and/or very highly plastic expansive clays obtained from two different construction sites in Ankara, especially on the swelling properties. However, this result is not valid for the agent fly ash which is an industrial waste. It is recommended to use fly ash with a calcium-based agent like lime or cement to improve such type of very highly plastic expansive clays in terms of swelling criteria since utilizing fly ash especially is quite advantageous from an economical and environmental points of view.
Keywords: expansive clay, lime, fly ash, treatment, swell potential, curing time, unconfined compressive strength, resilient modulus