İstanbul şehir surları Horasan harçları üzerine bir araştırma

Abstract

ÖZET Eski eserlerde harç ve sıva onarımı, bugün oldukça ciddi problemler ortaya koymaktadır. Binalarda işlevsel gereklilikleri yanı- sıra, estetik ve tarihi açıdan önem taşıyan harç ve sıvalar, restorasyon çalışmalarında da, önemli rol oynar. Bu araştırmada da, Osmanlı ve Bizans döneminden kalma birçok tarihi eserde kullanılmış olan hora san harcı incelenerek, gelecekte yapılacak onarım çalışmalarına yar dımcı olmak amaçlanmıştır. Geleneksel kireç harçlarının, sertleşme için hava ile sürekli temas halinde olma gerekliliği, suya direncinin olmaması gibi dezavan tajlarını ortadan kaldırmak üzere, çeşitli puzolanik malzemelerin harç karışımına katılımıyla hidrolik bağlayıcılar elde edilmiştir, tşte bu hid rolik bağlayıcılar grubunda yer alan horasan harcı, havaya ihtiyacı olmadan, su ile kimyasal reaksiyona girerek katılaşır. Dolayısıyla nem li ortamlarda kullanıldığında, daha iyi katılaşacak harç, suya karşı son derece dirençli olması sebebiyle, kireç harçlarından üstündür. Ancak, harç karışımında bulunan puzolanik malzemelerin, harç içinde puzolanik bir etki yapması çok uzun zaman alır. Çimentonun bulunma sıyla önemini yitirmiş olan bu harç, gücünü günümüze ulaştırdığı tari hi eserlerde kanıtlar. Horasan harcının, onarım çalışmalarında yerini çimento veya kireç harçlarına bırakması sonucu, son derece başarısız uygula malar yapılmıştır. Bunu önlemek amacıyla, bu çalışmada horasan har cının analizi yapılmış ve daha önceki araştırmalarda tespit edilmiş olan, harç karışım oranları verilmiştir. Laboratuvar çalışmalarında, İstanbul şehir surlarından Sulu- kulekapı, Edirnekapı ve Balat bölgelerinden alınan örnekler incelen miştir. Deneysel çalışmaların sonucunda, numunelerin çoğunluğu tuğ la kırığı, kuvars feldspat, biotit, muskovit gibi mineraller, volkanik nitelikteki taş parçaları, kum ve çakıl gibi taşınmış elemanlardan oluş tuğu anlaşılmıştır. Örneklerde saptanan ufalanmış, toz haline getiril miş tuğla kırıkları ile volkanik taş parçalarının ise, puzolanik malze me olarak kullanıldığı düşünülmüştür. Ayrıca, genel özellikleri bakımından Sulukulekapı ve Edir nekapı numuneleri benzer olmasına rağmen, Balat numunelerinin fark lı olduğu tespit edilmiştir. Buradan da Balat harçları ile diğerlerinin farklı dönemlerde uygulanmış olduğu sonucuna gidilmiştir.

SUMMARY A RESEARCH ABOUT KHORASAN MORTAR OF ISTANBUL CITY WALLS The fundamental purpose of a mortar is a bedding agent which makes a wall into an integral unit by taking up irregularities in the masonry units and the emors in positioning and thus events out the distribution of load in the wall. Many old masonry buildings with thick walls have mortars of only limited strength which are, nevertheless, perfectly satisfactory in this role. Generally, the mortar is also required to prevent infiltration of water from weathering effects. Increasingly, mortar is also being relied on to give the masonry structure flexural as well as compressive strength. Except their function, ancient mortars are valuable evidence on old masonry techniques. They provide information about the construction techniques of the relevant building period. In course of time, mortar which is the main subject of this research is deteriorated such as all parts of the historical building and needs repair. A general requirement for materials to be used in the repair of ancient structures is that, besides performing satisfactorily their mechanical function, at the moment of execution, they should not accelerate the deterioration rate of the adjacent, ancient material. Before embarking restoration adequate information is required about the original building materials. After that it is possible to apply, new techniques and matching materials which are suitable functionally and visually. It seems to be clear that importance of research before restoration and role of mortar in these is increasing day by day. In this study primarily, component of mortars, materials of them were researched and characteristics of modern and ancient mortars are compared. XIBinders, gypsum, lime, as an artificial cement, portland cement, as a natural cement pozzolanic cements were examined. Gypsum has been used as a binder since the time of Pyramids. It is usually found in rock formation in various parts of the world. The rock is crushed, ground and calcined which drives of about 75 percent of the combined water, forming the hemihydrate known as plaster of paris. If this product is mixed with water a chemical reaction takes place and the original rock structure is reformed. It has several advantages over traditional lime, including lower fuel requirements for manufacture, greater strength and more rapid setting time. It's major disadvantage is it's relatively high solubility. Lime was commonly used in the past as a constitute of masonry mortar. Limestones, including chalk, provide the raw materials for lime. The lime used is hydrated or slaked lime, produced by slaking quicklime, which was produced by burning limestone. In lime kilns slaked lime sets gradually by losing its water through evaporation and absorbing carbon dioxide, transforming from Ca(OH)2 to CaC03 or limestone. In praticular, lime mortars show a relatively high deformability, which allows them to accommodate some differential movements and a good ratios of large pores in their pore-size distribution which should favour movement of water vapour. Because of these advantages, there has been a tendency to specify the use of traditional lime mortars for masonry repairs. This in turn, has met with other difficulties. Lime mortar requires considerable skill for water addition which affects mechanical strength needs relatively dry ambient air and carbon dioxide for setting which is so slow. In ancient masonry which is frequently moist and had thick structural elements carbonation of the core is extremely slow and unreliable. Portland cement is on artificial cement obtained by intimately mixing together calcareous and argillaceous or other silica, alumina, iron oxide-bearing materials, burning them at a dinkering temperature and grinding the resulting clinker. Although, portland cement is a good hydraulic cement that hardens under water, the use of it in restorations should be restrained. Portland cement by being stronger then original material and having a greater thermal expansion coefficient which adds stresses on old masonry, plus during production having some alkali elements that will lead to the generation of water soluble salts and by having densly small pores that hinder water movement in masonry. All of these reasons, explain why portland cement shouldn't be used frequently in conservation. xuNatural cements are eminently hyraulic limes. The phenomenon of hydraulic set seems to have been appreciated first in Mediterranean countries under Roman influence, where there was sources of natural materials ejected from volcanoes. These were in the form of rocks such as tuff, trachyte and pumice or deposits of volcanic ash or earth such as pozzolana. Roman builders used bricks, tiles and pottery crushed to dust and ground iron slag as pozzolanic additives. Such materials contain, reactive silicates which in the presence of water, react with lime. In our country pozzolanic additives had been used with the name of khorasan. Khorasan is the name given to crushed and ground burnt clay as in bricks, roof tiles, and pottery. The khorasan mortar is a mortar made with khorasan and slaked lime sometimes with additions of sand. The setting of khorasan mortar is completely different from lime mortar. The hardening of khorasan mortar, however is based on the formation of water insoluble salts through an acid-base reaction. This reaction progresses at a very slow rate, but the presence of air or carbon dioxide is not needed, the aqueous media in the mortar is necessary for the reaction to proceed. In time, cost and availability problem of puzzolanic materials, restricted their use. Other sources had to be found to impart the necessary and desired properties to an often-weak lime and sand paste. Organic additives such as blood, egg, sugar, cheese, dung, gum arabic, kragacanth, animal glue, milky juice of figs, keratin, casein, rye dough, hogs, lard, curdled milk offered the mason a variety of solutions. The present study is dealing with the characterisation of khorasan mortar which had been used in Istanbul City Walls. Samples were taken from Balat, Edirnekapi and Sulukulekapi regions and analyzed by the following experiments. VISUAL ANALYSIS: Visual analysis were done to get some information about samples' physical properties such as colour, quality of compound and textural constituents. Results are given in Table A.l. COMPRESSIVE STRENGTH: Compressive strength of the samples are determined to evaluate the power of samples. Results are shown in Table A.2. WATER ABSORPTION: The water absorption coefficient was calculated from the experimental curve obtained by measuring the water absorbed by capillary action as a function of the square root of time. Results are reported in Table A.3-A.6 and Figure xuiA.11-A.14. WATER IMBIBITION: The water imbibition was determined by 24 hours immersion at atmospheric pressure and the water absorbed expressed as percentage of the dry mass of samples. Second water imbibition experiment which was with pressure was realized by boiling samples about 2 hours. The results are given in Table A.8 and Figure A.15-A.18. OTHER EXPERIMENTAL WORK ON PHYSICAL PROPERTIES: The spesific weight, dentistry of samples are measured and given in Table A.9-A.10. CHEMICAL ANALYSIS: Mortar samples were solved in hidrocloric acid to determine the proportions of mortar constituents. The basic principle of this analysis was first to dissolve the lime binder in acid, then to seperate the aggregate and the fines and thereby determine their proportion. Results are shown in Table A. 11, Figure A.19-A.26. Chemical analysis included the procedure aimed at finding losses an incineration, moisture content and percentage of carbonate content. The results of calcination analysis are given in Table A. 12. In the experiment programme of this study, first of all visual analysis of samples realized samples were seperated to different groups according to their colour and texture. Samples of Balat which have too much brick powder were in dark red. The colour of Suiukulekapi and Edirnekapi samples which have less brick in them were varying from pinky-red to whitish cream. Amount of brick which was in mortar effected the colour of it. To get better information about mortar after chemical analysis aggregates of the mortar were examined under binocular microscope. In samples crushed and ground brick were mostly used as aggregates and artificial puzzolanic materials. Sand, grit, clay, biotit, muscovite feldspath and piece of volcanic stones were the other materials found in mortar. Khorasan mortar has proved the strength of it by contributing buildings to survive for long years. In that time extrinsic forces already affected on mortar and deteriorated. Even though that compression strength, experiments were done to get some general ideas about samples' compressive strengths. According to the results of experiments, the compressive strength of Suiukulekapi samples are between 4-7 MPa Edirnekapi samples are xiv4-6 MPa and Balat samples are 1,5-2 MPa. It is shown that even though Sulukulekapi samples have little bit higher value than E dim e kapı, they have similar results but quite different from Balat samples. At it is known weathering, can destroy materials which have direct contact with it. When the compressive strength of samples and location of them are compared, it is found that, samples which are from face work of walls have lower strength than the core samples. The next experiment was made for obtaining the amounts of water absorption of samples. In the end of this experiment was obtained that Sulukulekapi samples water imbibition ratio is higher than the others. Balat samples which have the least water imbibition, the highest moisture absorption capacities, the highest water absorption coefficient, Edirnekapi samples have exactly opposite of these ratios. When Sulukulekapi samples were examined about moisture capacities has found that I-8A numbered sample which is from the northern part of stairs has the highest ratio 1-13 numbered sample which is from Theodosion time (taken from below ground level) has the least one. When water imbibition experiment supported by visual analysis, real density experiment it was found that there is a correlation between samples' porosimetry and water absorption. Mortar with larger pores has the highest water absorption coefficient but capacity of water imbibition changes by pore size distribution. Another remarkable point is role of samples' location. It is found that face work samples absorbed more water than internal samples. Chemical analysis have been done to seperate aggregate and fine in acid solution and thereby assess their ratios. They provided very important information about components of mortar. After lime of mortar dissolved, aggregates were seperated with the fine sieves according to their size and granulometric graphics were drawn. From here it is found that while Balat samples' aggregates which are from finely ground brick, have the largest grain sizes, Edirnekapi samples have the smallest grain sizes. In chemical analysis programme Sulukulekapi samples were examined more than the others. 1-10 and 1-7 numbered samples having similar minerals in them, gave similar granulometric graphic, seem to be belong to the same era. Because of having different proporties from the original mortars, it is better to accept they were used as repair mortars. Another interesting sample is by having most materials on 1000 [i fine sieve, sample 1-18 was paleologus repair. After examination under binocular microscobe it was determined that even though this ratio, it has the similar properties with 1-17, 1-19 xvbesides increasingly of their granulometric graphics similar among them but less than the others. Except these samples determined that south and north walls mortar samples (1-13, 1-14) which were below ground level with yellowish-white coloured, exhibited similar mineral composition and granulometry seem to be Theodosion. Calcination analysis assessed that, Sulukulekapi mortar samples were composed of about 14-43%, Edirnekapi samples 14-18%, Balat samples 20-48% calcium carbonate (CaC03). In Sulukulekapi samples which are from 78th tower archway (1-5) has the least amount (14%) of CaC03 but other samples which are from inventory of laboratory (I-a,b,c), infill of the gateway (1-1), 77th tower north wall (1-2), north part of stairs (1-8) and Theodosious time imply quite large amount (from 25% to 43%) of CaC03. Conceptually it is possible to say pinky-red and whitish-cream coloured Sulukulekapi samples mostly have approximately 20-25% CaC03. At the end of research it was found that samples which were used in this research mostly made of crushed or ground brick, quartz, feldspath, muscovite, biotite and some volcanic stones. It is realized that crushed or ground brick and piece of volcanic stones were added to the mixture as pozzolanic additives. Samples mostly have the general characteristics of khorasan mortar however it was found that similar but mainly different properties of mortars exhibited the probability that they were used in subsequent repairs. Finally, according to all analytical work it was assessed that Sulukulekapi and Edirnekapi mortars had similar, whereas Balat samples had quite different characteristics. xvi