The Effect of additives on the shear strength of brick masonry wall under biaxial loading

Tez (Doktora) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1990Thesis (Ph.D.) -- İstanbul Technical University, Institute of Science and Technology, 1990 Ülkemizde yığma yapılar kırsal kesimde yaygın biçimde kullanılmaktadır. Yığma yapılar üzerinde yapılan araştırmalar, iskelet taşıyıcı sistemlerle ilgili araştırmalara göre daha az sayıdadır. Duvarlar sadece düşey yük etkisinde kalmayıp, rüzgar, deprem veya başka nedenlerle yatay yük etkisine de maruz kalırlar, böylece düşey-yatay tesirlerin birlikte etkimesinde, duvarlarda iki eksenli yükleme hali meydana gelir. Tuğla duvara yatay bir yük uygulandığında harcın, tuğlaya göre dayanımının az olması ve tuğla ile harç arasındaki aderans köprü bağı yeterli olmadığından genellikle çatlaklar harçta oluşmakta, ayrılmalar tuğla ile harcın birleşim yerlerinden olmaktadır. Yapılan çalışmalarda tuğla duvarların iki eksenli yükleme altındaki davranışı deneysel ve analitik olarak incelenmiştir. Bu çalışmalarda tuğla duvarın yatay yüklere dayanım göstermesi için donatılı yapılması önerilmiş, harca aderans arttırıcı katkı maddeleri koyan çözüm yolu pek incelenmemiştir. Bu çalışmada harcın içine çeşitli katkı maddeleri konarak, tuğla ile harcın arasındaki aderansın iyileştirilmesine, tuğla duvarın çekme dayanımının arttırılmasına çalışılmış iki tip matematik model sonlu elemanlarla çözülerek, deney sonuçları kıyaslanmıştır. îlk aşamada 45 numune üzerinde çeşitli harç katkı maddeleri kullanılarak aderans ve çekme dayanımlarıyla ilgili ön deneyler yapılmıştır. Bunların sonucunda bazı katkı maddeleri araştırma dışı bırakılıp, 7 tip katkılı ve bir katkısız harçla 100 küçük numune derzlerine göre iki eksenli yükleme altında denenmiş, a- t ilişkisi eğrileri karşılaştırmalı olarak çizilmiştir. Kayma kırılmalarından, basınç kırılmasına geçişin 45 ilâ 60 arasında olduğu bulunmuştur. Bu nedenle a=30-45° arasında kayma dayanımı maksimum olan Parılat L20 katkı maddesi ve Embet hazır harcı seçilerek, duvar düzlemi içinde üstten uniform yayılı kenar yükü ve yandan yatay basınç yükü yüklenmiş, 28 numunede ct- t ilişkisi eğrileri, normal harçla karşılaştırmalı olarak çizilmiş, katkı maddesinin elemanın taşıma gücüne etkileri araştırılmış, düşey ve yatay deformasyonlar ölçülerek gerilme-birim deformasyon eğrileri çizilmiştir. Enerji yutma kapasitesini belirlemek amacı ile deprem kuvvetine benzetilebilen yatay yük etkisindeki duvar elemanın yatay deplasmanları ölçülerek yük-yatay deplasman eğrileri çizilmiş. Çeşitli katkı maddeleriyle örülen duvar numunelerinin enerji yutma kapasiteleri araştırılmıştır.Masonry structures are extensively utilized for various purposes in the rural regions of Turkey. This type of structure is vertically supported by walls which play an important role in overall strength of the structure. The research carried out on masonry structures is comparatively less than that on frame supported systems. The results of experimental work in this field show great diversity. It is not suprising for a brick wall to yield strength measurements widely differing from those obtained by theoretical calculations. The compressive load carrying capacity of brick wall varies between 30 % - 50 % of that of brick. The mortar used is a significant factor in the resultant compressive strength of wall. A brick wall is a composite structure and it is not a homogeneous or isotropic medium. It may be strong but is also brittle, its compressive performance being stronger than its performance under tension. Walls carry not only vertical loads but also horizontal loads due to wind, earthquake or other reasons. Under such loading, the walls are subjected to two-dimensional loading. In horizontal loading, the fractures occur at the mortar joints and the structure fails at the bond between mortar and bricks. This is due to the low bonding strength between the brick and the mortar. The performance of brick walls under two-dimensional loading has been studied both experimentally and theoretically. In these studies, various methods are proposed to increase the strength against horizontal loading, but the effects of adding bonding agents to the mortar has been relatively neglected. In this study, various additives are used in the mortar in order to increase the bond between mortar and brick and to improve the tensile strength of the wall. Variations caused by additives in a-r t curves of the wall have been investigated. A brick wall is a heterogeneous material. Its properties are not solely dependent on the properties of viii brick and mortar used in the construction. Other significant parameters are the quality of bricklaying, dimensions of joints, water absorbing capacity of bricks, consistency of mortar, overall dimensions of the wall, etc. These parameters are not always describable in terms of numerical quantities. Therefore, it is likely that theoretical analysis will be severely dependent on the assumptions related to these parameters. Under these circumstances it is thought that better and more useful results could be obtained by performing a large number of experiments. Mainly the experimental approach has been followed in this study. 2. EXPERIMENTAL STUDY In the initial stage of the study, 45 samples are tested with various mortar additives, investigating the bond and tensile strength of the samples. These tests constitute a set of preliminary experiments. As a result of the preliminary experiments, some of the additives have been excluded from the test program. (Then 7 additives have been used in 100 relatively small samples. A number of samples without additive are also included into the test program.) In this set of tests, o - t curves are obtained, in a comparative fashion, under 2-D loading. tga =*M "*p. " iKhj (Here a is the angle between the resultant of the applied load and the horizontal reference). The transition from shear failure to compression failure has been found between 450-60°. For this reason, a new set of tests have been devised where two mortar samples (Parxlat-L20 additive and Embet Ready Mortar) which showed the highest shear strength between 30°-45° are used. (In these tests, the loadings were chosen as a vertical, downwards, uniform edge load and horizontal in-plane compression load) On 28 samples, a- t. curves have been drawn, including the comparisons with normal mortar. The effects of the additives on the strength of the wall have been investigated. The vertical and horizontal deformations have been measured. The sketches illustrating the ultimate failure modes of all samples obtained from each experiment have been given to emphasize the physical conclusion of the experimental study. Earthquake is basically the release of the potential energy stored in the earth's crust, so it is necessary to build all structures to absorb such an energy. To determine the energy absorbing properties of the sample walls, presenting static earthquake loading an horizontal load has been applied in the tests. Load-horizontal ix deformation curves have. been drawn in order to Investigate the energy absorbing capacities of the walls built using various additives. 3. MEASUREMENTS AND RESULTS The first set of experiments were performed to investigate the type of additives to be used in the main sample wall tests. In these tests, sand (Riva) and cement (KPÇ 32 5) types were kept unchanged. The additives chosen, here were Sika-Latex, Retan-707, Plastizayr-BV. These additives were found to yield tension and pure shear stresses higher than normal mortar. On this evidence, these additives were selected for later experiments. The main result of these pilot experiments was found that especially Plastizayr-BV and Retan-707 provided higher strengths. The second set of experiments were performed with a mechanical set-up where 2-D load was applied to the small wall samples through concrete blocks situated at the diagonal corners of the sample as shown below. CONCRETE BLOCK PRESS THE WALL SAMPLE OF 'K' TYPE The additives used in these tests were Sika-Latex Plastizayr-BV, Parilat-L20 and Embet Ready Mortar. The angles of these loads with the horizontal level were a =23?,30°, 45°, 60°, 75°, 90°. The horizontal and vertical deformations were measured only at a=90° as a preparation for later tests. For deformation readings, the load was increased in steps of 1/10 th of failure load predicted from previous tests. For this purpose, a Hugenberger mechanical extension gauge was used at about the center portion of the sample with initial measurement distances being 10 cm. in both vertical and horizontal directions. These directions coincided with the directions of the horizontal and vertical mortar joints. The global failure of samples were found to be in either shear or compression. The shear failure is defined as the loss of bond at the joints between brick and mortar. The compression failure is defined as the cracking or crashing observed in the bricks. When a=23° to 45, the failure nearly always occurred in shear mode, and when a = 45 to 60 it was found to be the transition region from sehar to compression failure mode. A summary of the results is shown in the figure below. aC (kg/cm2) /25%^ norma!" ~l - i - i - r~ ro to _ oi "ray -t - i - r-r^r « w ro to roCf (kg/cm2) *0 o *?* m f._ The shear failure is due to the yield of the shear strength of the mortar, and the bond between the mortar and brick. Therefore, the a - t curves between a=o to 4 5 appears as a straight line. This line could be expressed as Tb = Tbo + ya where p is the coefficient of friction, between the mortar and brick, Lbo is the shear failure stress, is pure shear stress. (Embet Ready Mortar) (Parxlat - L20) (Plastizayr - BV) (Normal Mortar) (Sika - Latex) = 8.24 + \io = 6.69 + ya = 3.26 + ya = 2.06 + ua = 1.37 + \ia A general conclusion derived from the experiments described above was that the additives Parilat-L20 and Embet Ready Mortar produced joints highly resistant to shear loading. Because of high bond strength of these additives even at a = 23, instead of shear failure as expected, some cracking is also observed in the brick. Depending on the results of the first set of experiments, a second set of tests was devised. In these tests, the samples with larger dimensions were subjected to uniform vertical edge loads and singular horizontal loads by a XI rigid frame, manufactured using double, I section profiles. Shown at photograph below. Hidralic jack Teflon strip »; Extensometer Wood support Rigid frame I Profile Hydrolic jack Masonry wall sample The uniform vertical edge load and the horizontal load were both delivered by hyraulic jacks which were equipped with manometers for load per unit area readings. As the capacity of the manometers of the previously used jacks was exceeded, the samples with a= 90 loading were tested in a Amsler hydraulic press with a capacity of 500 tonnes. The deformation reading were taken in similar manner to the previous set of experiments with a mechanical extension gauge, with 1/500 mm accuracy. The initial measurement distance was increased to 20 cm due to larger dimensions (40x40 cm) of the wall samples. The readings were taken as averages of measurements at the middle of the front and back surfaces for horizontal deformations, and as averages of measurements taken at positions close to vertical edges for the vertical deformations. In these experiments, an additional measurement was taken as the horizontal displacement at level of the side load applied to the sample by again a mechanical extensometer gauge with 1/100 mm accuracy. The load was applied at steps of 1/10 th of the failure load predicted from the previous small sample tests. As decribed before, the cr _ x of straight lines were found as: curves in the form xb = 7.82 + 0.746a xb = 4.63 + 0.612a Tb = 2.65 + 0.735a (Embet Ready Mortar) (Parilat - L20) (Normal Mortar) xix These results, with the expection of slightly lower values obtained for Parxlat-L20, appear to be unchanged by the change in the dimensions of the sample walls. This is particularly true for the shear failure. The additional measurement of horizontal displacement under horizontal loading was used to indicate the energy absorbing capacity of the samples. For this purpose, load-displacement curves were drawn and the areas under the curves were evaluated to estimate the energy absorbing capacity. Embet Ready Mortar was also found to be advantageous in this respect. 4. ANALYTICAL STUDY For analytical study two finite element models were proposed. The material properties included into elastic anally ses, are determined from the tests. 4.1. FIRST MODEL Brick and mortar joints are modelled separately with individual and different properties, by using finite element method. It is assumed that the mortar as frame elements and the brick as quad elements. 4.2. SECOND MODEL It is assumed that brick and mortar together as composite quad members. Pfcr/3 is applied to specimen assuming that it behaves elastic ly up to this level. The measured horizontal displacements of two test specimens are compared with the computed results of the proposed model. 5. GENERAL CONCLUSIONS 1. Sample walls under 2-D in-plane loading, fail either in shear or in compression, depending on the ratio of the horizontal and vertical components of the load. 2. The ct-t curves are straight lines when the loads have angle of a.- 0 to 45° with the horizontal joints (the shear failure). These curves become parabolas when a>45 (The compression failure). 3. The transition from shear failure to compression failure is observed to be between a= 45°and 60° 4. The effects of additives on the shear behaviour of masonry brick walls are investigated and it is found that some of these additives significantly improve the shear performance of such walls. Xlll As it could be seen from the a - t curve equations, average value of Tbo is increased 2.95 times with Embet Ready Mortar and 1.76 times with Parilat-L20 compared with simple mortar values. This increase also effect the yield load of specimens positively. 5. If.the tensile strength of mortar specimens increases with an additive, such a mortar effects the shear strength of the wall in a positive (increasing) manner. 6. The horizontal load - displacement curves were studied under the assumption that the energy absorbing capacity is an indicator of earthquake resistance, In general, those additives that improve the shear behaviour were also found to increase the energy absorbing capacity of the sample walls. 7. The deformation measurements show that some additives increase the flexibility of mortar and hence, the sample wall behaviour changes accordingly. The deformation measurements show that some additives increase the ductility of the mortar and this affects the behaviour of the wall sample. 8. The changes in the dimensions of the sample walls are in general, observed not to affect the stress measurements in the shear loading mode (a = 0° to 45° ). 9. The proposed finite element model can be applied to brick-mortar combinations laid in any bond pattern, once the basic material parameters have been obtained from relatively simple tests in elastic region. The accuracy of the analyses will be influenced by the degree of variability in the material properties. Stress history, creep, and workmanship effects will also exert some influence in actual walls..DoktoraPh.D