Influence of aggregate volume concentration on the mechanical properties of concrete

Tez (Doktora) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1995Thesis (Ph.D.) -- İstanbul Technical University, Institute of Science and Technology, 1995Sunulan bu çalışmada normal ve hafif agregalı betonlarda agrega hacim konsantrasyonunun betonun kısa süreli elastik ve elastik olmayan mekanik davranışına etkisi araştırıldı. Üretilen betonlarda en büyük agrega boyutu, granülometri ve su/çimento oranı sabit tutularak agrega hacim konsantrasyonu değiştirildi. Disk yarma deneyleri yardımıyla betonların şekil değiştirme kapasiteleri ölçüldü ve agrega konsantrasyonunun bu dolaylı çekme halindeki şekil değiştirme kapasitesine etkisi incelendi. Basınç halinde tepe noktası öncesinde yükleme ve boşaltma yapılarak normal agregalı betonların gevreklik indisleri de bulundu ve bulunan değerlere agrega konsantrasyonundaki değişmelerin etkisi araştırıldı. Kırmataş agregalı betonlarda, agrega konsantrasyonunun zamana, bağlı davranışa etkisi, rötre ve sünme deneyleriyle incelendi. Sertleşmiş betonların elastisite modülleri iki fazlı bir kompozit malzeme modeli yardımıyla hesaplandı ve elde edilen sonuçların deneysel değerlere yakın olduğu bulundu. Agrega konsantrasyonundaki artışın çakıllı normal ve ponza taşı agregalı hafif betonlarda, süreksizlik sınırındaki Poisson oranını düşürdüğü, kırmataşlı normal betonlarda ise bu oranın bir minimumdan geçtikten sonra arttığı görüldü. Agrega hacim konsantrasyonundaki artışın hafif agregalı ve çakıl agregalı betonlarda süreksizlik ve çözülme sınırlan ile basınç dayanımını düşürdüğü, buna karşın kırmataş agregalılarda ise önce bir miktar azalttığı ve bir minimumdan geçtikten sonra arttırdığı görüldü. Ayrıca konsantrasyon artışı tüm betonların basınç dayanımındaki birim kısalmalarını ve kırılma-şekil değiştirme işlerini azalttığı gözlendi. Agrega konsantrasyonundaki artışın kırmataş ve çakıl agregalı betonların yarma-çekme dayanımını arttırdığı, buna karşın hafif agregalı olanlannkini düşürdüğü, çekme şekil değiştirme kapasitesini ise azalttığı bulundu. Normal agregalı betonlarda agrega hacim konsantrasyonu arttıkça gevreklik indislerinin başlangıçta azaldığı ve bir minimumdan geçtikten sonra arttığı saptandı. Gevreklik indisi değerlerinin silindir basınç mukavemetlerindeki artma ile belirgin biçimde arttığı görüldü.Concretes with compressive strength exceeding 80 N/mm2 are now commonly used in the construction of some important structures such as long span bridges, dams, nuclear reactors and off-shore structures. in some countries, it is now possible to produce concrete with a compressive strength of 115 N/mm2 in commercial applications. Since concrete is weak in tension, its tensile strength is frequently neglected in the conventional design of reinforced concrete structures. in recent years, research on concrete have primarily concentrated on increasing its compressive strength, however, more infonnation needed for many aspects of mechanical properties such as fracture behavior and tensile properties. The infonnation about tensile properties of concrete is required for calculating the response of a structure, especially for determining the risk of cracking. The wide use of modern computer-aided analysis in design and use of concrete for special structures mentioned above have led to growing interests in cracking behavior of concrete. it is expected that in the next century more sophisticated approaches will be used in design and, as a result, more research will be recraired on concrete for a better understanding of its material behavior. in some loading configurations, such as uniaxial tension, splitting and biaxial loading, opening mode cracks are typical. Concrete is much more sensitive to tensile failure than to shear failure and it is impossible to create a püre shear stress distribution in a test specimen without introducing tension. it has been suggested that initiation of cracking is more dependent on strain than on stress. When a tensile strain is hıduced in concrete, if its value is greater than the tensile strain capacity of the concrete, a crack will be initiated. The tensile strain capacity is defined as the strain at the onset of cracking ör at the limit of proportionality between stress and strain in tension. it is not a constant value, but depends on the experimental technique used, on the type, size and grading of aggregate, on the gauge length, and on the water/cement ratio, curing conditions, age of concrete and loading rate. Available test results in literatüre show that there is good correlation between tensile strain capacity and strength of concrete for various curing conditions and ages. it is known that, the short-term tensile strain capacity of concrete can be predicted if the modulus of elasticity and tensile strength is known. in recent years, detailed research on the termal strain capacity of concrete was done. it is shown that the thermal strain capacity of lightweight concrete is greater than that of flint gravel concrete of similar workability and strength. There is a good relation between the thermal strain capacity and "strength/modulus of elasticity"ratio. Early investigations show that the maximum strain at peak stress was about 100xlO"6. Direct tensile behavior of concrete under long-or short-term loading was investigated by various researchers. Engineers, however, need more information on the tensile strain capacity of a wide range of concretes for design and quality control purposes. The main object of this work was to hıvestigate the influence of agregate volume concentration on the elastic and inelastic behaviour of normal and lightweight agregate concretes under uniaxial short-term compression loading. The influence of agregate concrentration on the splitting-tensile behaviour and the strain capacity of concrete were also investigated. Concrete was considered as a two phase composite material consisting of hardened cement matrix and agregate particles dispersed in it. Since the uniaxial tension test is difficult to perform, the split-tension test was prefered to measure the tensile strain capacity. in addition the brittleness index of normal weight aggregate concrete and also the creep and shrinkage of concrete (only crushed stone) were investigated. Finally the modulus of elasticity of concretes measured experimentally and compered with the models estimated by using two phase composite material models. The concretes produced in this study are as follows: 1) Eleven different aggregate volumes such as O %, 17 %, 20 %, 34 %, 40 %, 50 %, 51 %, 60 %, 68 %, 80 % with crushed limestone series, 2) Six different agregate volumes such as O %, 20 %, 40 %, 50 %, 60 %, 80 % with lightweight agregate (pumice) series, 3) Six different agregate volumes such as 20 %, 40 %, 50 %, 60 %, 80 % with gravel agregate series, 4) Eight additional agregate volumes such asO%,8%,16%, 24%, 32%, 40 %, 48 %, 56 % with lightweight agregate (pumice) series without sand (0-2 mm), 5) Eight additional agregate volumes such as O %, 8 %, 16 %, 24 %, 32 %, 40 %, 48 %, 56 % with crushed limestone series without sand (0-2 mm), The aggregate volume concentrations have been changed in the range of O %- 80 %. The aggregate grading of concrete, water/cement ratio and the maximum partide size for ali concrete series have been kept constant. The water/cement ratio and the maximum size of aggregate were 0,28 and 20.0 mm, respectively. Ali mechanical tests were carried out on the standard cylinder specimens, the height=300 mm and the diameter= 150 mm, at the age of 180 days, but only for concretes without sand (0-2 mm), at 28 days age. Splitting-tensile tests were carried out on disk specimens which were prepared from the standard cylinder as the height=150 mm and the diameter=150 mm. Test results obtained are evaluated in terms of following properties: Fresh concrete properties, elastic and inelastic properties, fracture energy, under compression splitting tensile strength, tensile strain capacity, ultrasonic pulse velocity, creep, shrinkage, the brittleness index and the modulus of elasticity. The experimentally measured values and the predicted values using the composite model are compared. This thesis consists of five parts: in the first part, an üıtroduction is given containing the objective of the investigation, relevant general information and releated definitions. Also in this part the aim and scope of the present work are given. The second part is devoted to the experimental studies. The materials used, the principles assumed, the mix compositions, the methods of mixing and curing, the types of loading and the methods employed in the testing and measurements made are described. in the third part, the experimental results are presented. The experimental results are discussed and evaluated in the fourth part. in the fifth part the conclusions are summarized and suggestions for further study are given. The results obtained in the experünental work can be outlined as follows: 1) Results related to the elastic behaviour: i) Agregate concentration has an important effect on the modulus of elasticity, such that, when it increases the modulus of elasticity of normal weight agregate concrete increases, but the modulus of the elasticity of lightweight aggregate concrete decreases. ü) in concretes produced with limestone aggregates, poisson ratio (at the discontinuity limit) has a minimum around a concentration of 0.425 m3/m3, and for the same type of aggregate type but without sand series, it has a minimum around concentration of 0.270 m3/m3. in the siliceous gravel aggregate concetes and pumice lightweight aggregate concretes, poisson ratio decreases with hıcrease in aggregate volume in the mix. üi) For the purpose of modelling, the modulus of elasticity of concrete can be estimated from the moduli of aggregate and cement paste phases by using composite materials models. The predicted values showed reasonable approximations to the experimental results.. in the limestone aggregate concrete series (with fine ör no fine, 0-2 mm), Bache-Nepper-Christensen, Popovics, Hashin, Hansen, Counto, Hirsch-Dougill models gave good results. And also in the siliceous gravel aggregate concrete series with sand (0-2 mm) having modulus of elasticity below 40000 N/mm2, Bache Nepper-Christensen, Popovics, Hashin-Hansen, Counto, Hirsch- Dougill, Reuss and Voigt models gave good results.. In the pumice lightweight aggregate concrete series without sand (0-2 mm), Hashin-Hansen, Counto models gave good results. 2) The results related to inelastic behaviour: i) As the aggregate volume concentration of the pumice lightweight aggregate and siliceous gravel aggregate concretes increse the discontinuity limits decrease. However, in concretes with limestone aggregates, the discontinuity limit has a minimum value around an aggregate concentration of 0,51 m3/m3 and it has a minimum around 0,24 m3/m3 for concretes without sand (0-2 mm). ii) As the aggregate volume concentration of pumice lightweight aggregate and siliceous gravel aggregate concretes increase their, loosening limits (the onset of unstable crack propagation) decrease. In concretes with limestone aggregates, the loosening limit has a minimum value around the aggregate volume concentration of 0,34-0,40 m3/m3 iii) The compressive strengths of both the pumice lightweight and siliceous gravel aggregates series decrease, when the aggregate volume concentration increases. For limestone aggregate concrete series, however, the compressive strength first decreases as the aggregate concentration increases, then passes through a minimum point at an aggregate concentration of 0,20 m3/m3 and then increases. iv) The strains of all series at the maximum stresses decrease as the aggregate volume concentration increases. 3) The results related to the toughness and the relative toughness. i) As the aggregate volume concentrations of all concrete series increase the toughness of concretes decreases. ii) The relative toughness of concrete series does not change seriously, as the aggregate volume concentration increases. 4) The experimental results related to the splitting tensile strength and the tensile strain capacity: i) in normal aggregate concretes, as the aggregate volume in the mixes increases the splitting-tensile strength also increases. This increase is significantly higher in limestone aggregate concretes. On the other hand, in pumice lightweight aggregate concretes, as the volume of aggregate increases the splitting-tensile strength decreases after the aggregate concentration of 0,4 m3/m3 for concretes with sand (0-2 mm) aggregates and after 0,20 m3/m3 for those without sand (0-2 mm). ü) The tensile strain capacity is independent on the type and texture of aggregates within the limits of this work. iii) The tensile strain capacity as defined and measured in this work is strongly dependent on the volume of aggregate in the mix and it decreases with increase in aggregate concentration. 5) The results related to the ultrasonic pulse velocity: in normal weight aggregate concretes, as the aggregate volume concentration increases the ultrasonic pulse velocity also increases, but for the pumice lightweight concretes it decreases. 6) The results related to the brittleness of concretes under the uniaxial compression tests: i) in normal aggregate concretes, the brittleness index has a minimum value around the 0,30 m3/m3 aggregate volume concentration. ü) Within the limits of this work, the brittleness index does not vary for the normal weight aggregate concretes with the compressive strengths between 20 N/mm2 to 50 N/mm2, however after the value of 50 N/mm2 the brittleness index increases faster. Therefore, the value of 50 N/mm2 can be taken as a threshold strength for the typical brittleness of the aggregate volume concentration. 7) Proposals for further studies. i) Studies must be performed with different origin of aggregates other than those used this study, at different ages than those chosen in this study, and at higher water/cement ratios other than 0,28. ü) it may be interesting to investigate this problem in the case of flexural tensile loading using displacement controlled close-loop testing technigues. These kind of studies may be extended by calculating the fracture energies that includes the descending part of the load-displacement curve. Also it is reasonable to think that this problem may give more interesting results in case of biaxial loading. it would be beneficial to investigate with the brittlenes index in flexure. By using different types of aggregates, the results may be obtained in a wide range.DoktoraPh.D