EXPERIMENTAL ANALYSIS OF REINFORCED CONCRETE COLUMNS STRENGTHENED WITH STEEL TUBES

Transcription

1 The 5 th PSU-UNS International Conference on Engineering and 172 Technology (ICET-211), Phuket, May 2-3, 211 Prince of Songkla University, Faculty of Engineering Hat Yai, Songkhla, Thailand 9112 EXPERIMENTAL ANALYSIS OF REINFORCED CONCRETE COLUMNS STRENGTHENED WITH STEEL TUBES Ljubomir Vlajić 1, Miroslav Bešević 2*, Aleksandar Landović 2, Danijel Kukaras 2 1 Transortation institute CIP Belgrade, Nemanjina 6/IV, Serbia 2 University of Novi Sad, Faculty of Civil Engineering, Subotica, Serbia *Authors to corresondence should be addressed via miroslav.besevic@gmail.com Abstract: The toic of this aer is exerimental analysis of axially comressed reinforced concrete columns strengthened retrofitted with steel tubes. Reinforced concrete columns (squared section) were laced inside steel tubes, while free saces between the column and the tube were filled with different quality concrete mixtures. Key Words: Reinforced concrete column, strengthening, comosite section 1. INTRODUCTION Presently, domestically as well as worldwide, there is significant number of structures that are deteriorated, their usage has been altered or they have been damaged through different causes. Also, over the time, esecially for bridges, degradation rocesses of reinforced concrete (RC) elements caused by increased traffic load took lace, environment is becoming more aggressive due of industrialization and inadequate ecological rotection. Increased demand by new seismic codes caused some older structures to have inadequate seismic resistance. In extreme instances, it is necessary to strengthen a structure even during construction due to certain oversights within design and/or construction. Recently it's not uncommon to encounter a change of design reconditions during construction resulting in addition of one or two stories on to of already finished underground or ground levels of the structure. Restructuring, adatation and building extension often demands overall number of stories to be increased what, in turn, increases axial forces within RC columns. Figure 1 shows some tyical examles of damages to reinforced concrete columns. All of these structures demand some sort of strengthening for what, in case of columns, it is esecially convenient to lace steel tubes over the whole height of columns with emty sace being filled u with fine-grain concrete or secial concrete mixtures Figure 1. Examles of damaged columns Column, strengthen in this manner, has sufficient load bearing caacity for extra loads and its ultimate load bearing caacity is defined by the level of engagement of steel tube and the concrete infill or, in other words, the level of comosite action of the newly formed comosite cross section that consists of three different materials [1]. In rincile, strengthening is done by utilizing materials of similar or higher modulus of elasticity then the one of the existing cross section. Steel, as a material has modulus of elasticity that is over six times then the one of concrete and it reresents almost ideal material for strengthening - rehabilitating of concrete and reinforced concrete structures. Steel is characterized by high tension strength and ductility while concrete is characterized by high comression strength and rigidity. Combination of these two materials within one cross section creates comosite cross sections. This way advantage of both materials is utilized and structural elements have sufficient strength, ductility and rigidity. Comosite columns made from steel tubes filled with concrete reresent one of the first tyes of comosite

2 173 structures. Outer shell or steel tube enables that, due to couling effect with concrete, a hoo stress state forms what increases significantly the comosite action and load bearing caacity. Hoo stress effects cause biaxial stress state within steel and triaxial stress state within concrete core, while the concrete core itself local buckling of steel tube inwards. If the loading acts simultaneously on the steel tube and on the concrete core, at moderate load levels of.3.5f, different rates of exansion sideways of steel and concrete can occur due to different Poisson coefficient. In that case steel will have larger strains than concrete what will, in turn, decrease the comosite action of the cross section. σ 1,s σ 1,s Figure 1. Stress state in analyzed column σ 1,b σ 1,b The aim of resented research was that it, through exerimental - theoretical analysis and numerical simulation, rovides more reliable insight into behavior of axially comressed RC columns, or columns that have been strengthened by steel tubes. In other words, the aim was to define stress-strain resonse for incrementally increasing force until the failure of the axially comressed RC column strengthened by steel tube. Additional aim of the research was to define some recommendations for ossible use of these strengthened columns for wider alications within the construction industry, i.e. recommendations for aroriate rehabilitation of columns for "on site" conditions. 2. EXAMPLE OF STRENGTHENED COLUMN Tyical instance when strengthening of reinforced concrete columns is needed is when, due to construction of caitals middle columns are overloaded, esecially within ground level and basement stories. An increased load aears as a result of increased stiffness of the slab on account of the caital what causes a significant redistribution of the load towards the column beneath. Tyical examle of middle columns being overloaded and therefore with insufficient load bearing caacity, due to incomlete statical analysis, is shown in Figure 2. Efficiency of strengthening method of reinforced concrete (RC) columns with steel tubes is based on the effects of biaxial and triaxial stress states within base materials of the column. This rocedure was successfully introduced for rehabilitation and strengthening of several significant structures such as: "Medifarm" [1], Business tower "Ušće" in Belgrade, structure of the rinting house "Otimum" in Smederevska Palanka, Building "B1" of the Block 4 in New Belgrade, all conducted under suervision of rof. Ljubomir Vlajić, PhD. Figures 2. and 3. shows strengthening of object "B1". All of the mentioned objects are in Serbia Euroe. Figure 2. shows RC column with dimension of 4/4cm with caital Bv=2.5m just before strengthening with reviously described method. The motivation for strengthening was in the increased axial forces of the inner columns due to increased slab stiffness at the locations of caitals and small dimensions of the outer columns. Concrete quality was also one of the reasons that trigged a demand for strengthening. Column strengthening was erformed with tubes formed by cold forming of steel sheets with thickness of t=6mm. Steel shell was comleted with welding of four segments rolled into semicylinder shell shae (Figure 2.a). Finegrain concrete filling was erformed thorough holes at the to of the tubes, while comlete filling was conducted through hole drilled from the uer slab. Filling concrete mixture was "Polimag HK-8" reair two-comonent mortar roduced by Chemical factory "Prvi maj" Čačak. Main characteristic of Polimag HK grou of mortars is absence of shrinkage over the time, excellent comacting, good adhesiveness for concrete and steel and high mechanical roerties. It is also characterized by high curing rate with which comressive strengths of over 3 MPa were achieved high toughness and watertighness. Figure 2.b shows a steel sacer detail (sheet), which enabled efficient centric lacement of concrete column within the tube. Figure 2. Column strengthening with steel tube Additional convenience of this strengthening method is in the visual aearance of the columns, i.e. increased load bearing caacity is achieved without significant increase of the columns dimensions. Figure 2.a shows that the diameter of the steel tube is barely larger than the length of the column's cross section diagonal, so that the core column aears not to be significantly thicker, what is a result of the fact that square cross section column is rarely observed frontally. Figure 3. Final shae of the columns after strengthening

3 EXPERIMENTAL RESEARCH Exerimental analysis of this roblem was erformed on a basic model of reinforced concrete column with square cross section which was strengthened with steel tube. The testing was conducted with centrically alied load on columns constant cross section and hinges on both ends. Exerimental research included nine short reinforced concrete columns divided into three grous of three samles. First grou consisted of control samles of reinforced concrete columns without strengthening. Second grou consisted of three samles of reinforced concrete columns strengthened with steel tubes. Strengthening was conducted so that a reinforced concrete column was laced into a steel tube while remaining sace was filled with concrete mixture of the same quality as the core reinforced concrete column. Third grou of samles consisted from columns strengthened with steel tubes similar to the ones in the second grou only the filling concrete mixture consisted of secial high quality mortar Polimag HK-8 [3] and [2]. Figure 4. Secimens before exeriment Tested samles of axially comressed reinforced concrete columns had square cross section with dimensions 1/1 cm, height of 85 cm. Columns were made of concrete f c =3MPa, with module of elasticity E c =3Ga. Main reinforcements were ±2Ø5mm and stirrus of Ø4mm saced 3cm at the to and the bottom 2cm of the column length and saced 6cm in the middle. Steel quality was f y =5MPa. Strengthening of the columns were achieved with circular welded tube with outer diameter of D=159mm and wall thickness of t=2mm. Yield strength of the steel taken from steel tube was exerimentally determined to be f y =25Ma with module of elasticity E s =27Ga. Infill concrete had same characteristics as column core concrete, while Polimag characteristics were f =7MPa with module of elasticity E =32GPa. Tubes were formed with cold forming rocess and low-carbon welding rocess. Ratio of outer diameter and wall thickness was D/t=79.5 and it was chosen so that limit conditions for circular tubes filled with concrete are met as defined by: Euro code EC-4 [7], ACI [5] and AISC 36-8 [6]. Figure 5. Tested column samles a) control samle, b) samle filled with concrete, c) samle filled with Polimag HK-8 These dimensions were selected so that they corresond to real structure with a ratio 1:3,3. Model reresented a real RC structure model with low slenderness value (λ=29), dimension 33/33 cm, height of 28 cm and reinforcements consisting of ±2RØ19mm and stirrus of Ø1mm saced 1cm and 2cm, strengthened rehabilitated with steel tube with outer diameter of D=525mm and wall thickens of t=6.6mm. Considering that this exeriment was reared while resecting exact geometric similarity, theoretical - exerimental model analysis based on equality of Hookes constants, "design" and "rediction" equations was not erformed because direct analysis of exerimental results could be done as if the samles had realistic dimensions. For all 9 samles acquired and analyzed results included: changes in the stress and strain state, ultimate load bearing caacity, shae of the global deformations at failure, load carrying engagement of each material within comosite cross section. Local deformations (strains) were measured at the middle of the column's height. Figure 6. To of the comosite columns after exeriment Columns of the control grou showed linear behavior during whole loading rocess until failure (Figure 8). This behavior corresonds fully to one described in other literature, roving that axially comressed columns and high grade concrete columns show that their stress/strain diagram does not deviate much from the straight line.

4 P/Pu SP1 Figure 7. Crushing of the concrete core Beton/Concrete Čelik/Steel - V Čelik/Steel - H Deformation wise, the behavior of the comosite columns is almost identical and it is largely or comletely insensitive to the concrete fill quality. U to load level of 1/4Pmax a full comosite action so observed and behavior comlies with Bernoulli's hyothesis. Above this limit a deviation of strains ε1 is observed that was registered on RC column and on the steel tube, what brings us to a conclusion that sliding between materials within comosite cross section occurred. Comarison of strains shows that comosite columns had somewhat larger strains relative to RC column what is to exect regarding higher ductility of comosite cross sections. Furthermore, columns reared with Polimag have a larger strain caacity than ones reared with fine-grain concrete. Ration of main strains of the steel tube remained constant all the way u to failure and stresses within middle height cross section do not reach a steel yielding limit Concrete H P/Pu Concrete Beton V ε[1-6 mm/mm] -2 Beton H -25 Figure 8. Characteristic deformation resonse of the RC column Labels Concrete V and Concrete H reresents vertical (axial) and horizontal (lateral) strain registered on concrete, while Steel H reresent horizontal (hoo) strain registered on steel. All strains were measured at the middle of the column's height on all column models ε[1-6 mm/mm] Figure 9. Characteristic deformation resonse of the RC column strengthened by steel tube and Polimag HK-8 fill On Figure 8 and 9 value of the axial force in single load increment is reresented by P, while Pu is registered bearing caacity of column. Stress state analysis for each samle was conducted at cross section ositioned at middle height at loads of P=.5Pu, that is equivalent to maximum exloitation load. For evaluation of stress state within steel tube lane stress state assumtions were used. Measured strains were used then to determine stress state and comression stresses were registered for the longitudinal axes of the column, σ1, while tangential stresses, σ2, were tension stresses. (Figure 1.) Obtained stressed yield a conclusion that certain load distribution aeared between the tube and the core column as well as hoo stress effect within the tube itself. Transfer of load between materials within the comosite cross section was a result of friction forces at the adjacent surfaces of the cross section, i.e. steel and concrete. Table 1. Comarison of ultimate load forces for tested columns Column samle Mean value. P u [kn] % Control samle S Concrete infill SB Polimag infill SP Ultimate load forces, given in a Table 1, show that reinforced concrete columns strengthened with steel tubes and Polimag infill have aroximately 2.8 times higher load bearing caacity, while columns with concrete infill have aroximately 2.7 times higher load bearing caacity. Comarison of load bearing caacity of comosite columns shows that the difference is small, ractically nonexistent. Detail account of shown research is given within reference [3], while numerical analysis of this roblem is resented in reference [4].

5 CONCLUSIONS Exerimental results largely deend on assumtions made rior to design and forming of test samles what imlies that conclusions refer to exactly defined boundaries. According to results obtained through exerimental-theoretical analysis of samles following conclusion are made: RC columns from the control grou have almost linear stress/strain relationshi all the way to the failure. Columns strengthened with steel tubes show considerably more ductile behavior and are caable of withstanding larger deformations relative to classical RC columns. Strains observed at RC column and steel tube are almost identical until the load level of P=(.25.3)Pu, what means that comosite action of the cross section is reserved until this load level. (This is usually within the range of exloitation load levels). Strain diagram analysis revealed that whole cross section is engaged for load carrying. Transfer of loads between individual materials is done through friction of adjacent materials. (Steel tube and infill). At exloitation load it was clearly observed that hoo stress effects occur, i.e. biaxial stress sate in the steel tube aears and triaxial stress state aears within the core concrete column. Ration of main strains within the steel tube remains constant all the way u to the failure in the middle height cross section of the column and the strains do not reach a yielding limit. Force at which failure of the column occurs is equal to the force that crushes the core RC column within the steel tube. Ultimate bearing caacity of the columns strengthened by the steel tubes is increased by aroximately relative to the ultimate bearing caacity of the non-strengthened columns. Values of ultimate limit forces are ractically identical for all comosite columns and it was not observed that this force directly deends on the comressive strength of the infill. Failure of the control RC columns occurred as a result of cracking and crushing of the concrete at the location of load alication, while failure of comosite columns occurred as a result of the combination of concrete crushing and local buckling of the steel tube at the location of the load alication at the to of the column. 5. REFERENCES [1] Vlajić Lj., Kovačević T.: Investigation of comosite action efects of concrete-exmal-steel in case of axialy comressed columns, SDGK Symosium '89, Yugoslavia, 1989, [2] Vlajić Lj., Landović A.: Analysis of methods for strengthening reinforced concrete columns couled with steel tubes, 13th Congress of Serbian society of structural engineer, Zlatibor-Cigota, Serbia, 21, [3] Landović A.: Exerimental-theoretical model analysis of osibility for strengthening axialy comressed RC columns with steel tubes, Master s Thesis, Faculty of civil engineering, Subotica, Serbia, 21,. 11. [4] Vlajić Lj., Bešević M., Landović A., Kukaras D.: Numerical analysis of steel-concrete comosite columns under axial load, Serbian journal Izgradnja 64 (9-1), 21, [5] ACI-318M-8 Building Code Requirements for Structural Concrete And Commentary, ACI Committee 318, USA, 28, [6] ANSI/AISC 36-5 Secification for Structural Steel Buildings 25, American Institute Of Steel Construction, Fifth Printing, [7] Eurocode 4, EN :24: Design of comosite steel and concrete structures Part 1-1, Euroean Committee for Standardization, 24,. 225.