APPLICATION OF EXPANSIVE AGENT (EA) TO PRODUCE POST TENSIONING FORCE IN FRP JACKETS FOR LATERAL RETROFITTING OF RC COLUMNS

Transcription

1 CD APPLICATION OF EXPANSIVE AGENT (EA) TO PRODUCE POST TENSIONING FORCE IN FRP JACKETS FOR LATERAL RETROFITTING OF RC COLUMNS A.A.Mortazavi 1, M.Jalal 2, M.M.Khodaparast 3 1. Associate professor, Power and Water University of Technology, Tehran, Iran 2. M.Sc. Student, Dept of civil Engg. School of Engg Razi University, Kermanshah, Iran 3. B.Sc., Power and Water University of Technology, Tehran, Iran ABSTRACT Repair and strengthening of RC structures using FRP materials is one of the new and effective techniques ever used. With this respect, strengthening of structural elements by external lateral confinement can lead to increased strength and ductility. In this way, using an efficient and optimized method for lateral confinement is of great importance. A novel, economic and simple technique for the repair and strengthening of RC members by means of expansive agent(ea) to produce post tensioning force, has been proven to perform well following experimental work undertaken at the University of Sheffield. The aim of the technique is to ensure the enhancement of the member strength and ductility by localized strengthening. The above technique is equivalent to increasing the effectiveness of the composite confinement and it becomes possible to strengthen large columns with smaller amounts of composites, which are utilized earlier at much higher strengths. In addition, the level of axial strain achieved at failure is improved significantly. The paper will present details of experimental work with different types of confining material (glass and carbon), amounts of reinforcement and levels of initial prestressing. Keywords: strengthening, expansive agent, post tensioning, FRP jacket 1. INTRODUCTION Since the 1995 Kobe earthquake in Japan, composites have started being used for the repair and strengthening of columns against seismic actions. The composites are applied as external lateral reinforcement and are often used to prevent shear and anchorage/splicing failures, which can result in the enhancement of the ductility of RC elements. The Japanese philosophy on earthquake resistant design aims to achieve ductility through low ratios of reinforcement and low steel strengths. This results in very large sections, which as a result require little shear reinforcement. However, during seismic violent load reversals, the shear demand can be higher than estimated, for example as a result of vertical accelerations. In addition, splicing of reinforcement causes additional bursting forces, which are difficult to contain with nominal shear

2 600 / Application of Expansive Agent (EA) To Produce. reinforcement. External lateral confinement can address both of the above problems, in addition to providing other benefits. In Europe and New Zealand, the seismic philosophy for achieving ductility is different, relying more on increasing the non-linear concrete stains through concrete confinement [3]. The advantage of this philosophy is that apart from smaller crosssections, a significantly higher amount of lateral reinforcement is required. This lateral reinforcement, which is there to confine the concrete, is also beneficial in resisting additional shear and preventing splice and anchorage failures. The enhancement of concrete ductility by confinement is central to the principles of Eurocode 8. However, researchers dealing with FRP confinement do not always consult the huge wealth of published work, which is derived from the earthquake engineering research. In addition, there is a fundamental difference between mild steel confinement and high strength composite materials. Mild steel reinforcement attains its yield strength (around 80% of its ultimate strength) at a strain of around 0.002, whilst composites fracture at strains ranging from Unconfined concrete crushes when the lateral strain is at best This means, that for low levels of confinement, steel is relatively well utilised (around 50%), whilst composites are at best utilised at 7% of their capacities. 2. MATERIALS The properties of the fibres and resin used for confinement are shown in tables 1 and 2. Where: E frp Young s modulus of elasticity f frpu ε frpu Ultimate tensile strength and elongation of pultruded laminate T GM Glass transition temperature Volumetric fibre content V f+ Fibre GFRP AR CFRP 240 Table 1: Physical and Mechanical properties of Fibre sheets [6] E frp f frpu ε frpu Thickness Density T GM V f Composition GPa MPa % mm gr/cm 3 C o % Bi- Directional Uni-Direction Table 2: Physical and Mechanical properties of Epoxy Plus Structural Adhesive [6] Epoxy Colour Strength E m flexural Density T GK Coverage/ Manufacturer MPa GPa kg/litre C o Thickness U.K. Adhesive Mid grey mm SBD Primer Translucent m 2 /l SBD The next important material required for pre-stressing is the Expansive Agent (EA). It is supplied as a powder and the colour is grey when mixed with water [7] This material is normally used for concrete demolition and is placed in pre-drilled

3 3 rd International Conference on Concrete & Development / 601 holes. Figure 1 illustrates the relation between expansion pressure and reaction time for different hole diameters (when used for demolition) over a period of four days. Figure 1 also shows the EA when mixed with water (20% - 23%). In this experimental work the EA was mixed with cement in different proportions. Figure 1. Properties of Expansive Agent (EA) 3. THE EFFECT OF CONFINEMENT STIFFNESS In order to investigate the effect of confinement stiffness, the EG was confined directly by metal tubes. The tubes were selected to apply different confinement stiffness levels with different materials (steel or copper), thicknesses, diameters as shown in Figure 2 and different EA ratios. The ratio of length (L) to inner diameter (D) of the tubes was set to be around 10. It is obvious that, choosing different thicknesses of confinement materials (t), elastic modulus of the tubes (E) and radius of tubes or jacket (r), can change the confining stiffness (CS) as defined by 2E t / ID. Figure 2. Metal tubes filled with EG

4 602 / Application of Expansive Agent (EA) To Produce. Thirty six metal tubes were filled and tested with EG levels 5%, 10% and 20% of EA. The EG consisted of Betonamit (EA), 50% cement, the rest of mix ratio completed by sand (was variable) and appropriate water (between 17% to 20% of the weight of mix). Various methods of sealing the tubes were tried during casting to prevent any leakage due to the high pressure created by the EG. Due to the possibility of welding on the black steel, one end was fixed while the other end remained open until the expansive grout (EG) was poured into the tube. The end was then sealed by fastening a screw. Due to the characteristics of the metals, welding of copper and annealed steel tubes was not a feasible solution. Two solid plugs were therefore fitted with a located pin designed. In addition, a layer of silicon seal was applied to the surfaces of the plugs to prevent any leakages. The pipes were instrumented with 3 strain gauges placed at the mid length and equally spaced along the circumference. Two gauges were used to measure lateral strain and one to measure the axial strain. Since the strain gauge had a length of 15 mm, only two strain gauges, which were fixed one laterally and the other one axially were used for the smaller copper and black steel pipes. 4. EFFECT OF VOLUME OF EG This sub-phase investigated the effect of 3 different amounts of expansive grout (EG) with the same EA ratio (20%) on steel tube cylinders having the same confining stiffness. All tubes were instrumented on the outside with three surface strain gauges to measure lateral strain and one strain gauge to measure axial strain. The first sample (S1) was filled with expansive material without any core in the middle. The second sample (S2) had a concrete core with a diameter of 50 mm placed in the middle. Additional strain gauges were placed at the mid height of the concrete core to monitor lateral and axial strain. The third sample (S3) was similar to S2, but the concrete core had a diameter of 66 mm. Figure 3 shows samples S1, S2, and S3. To avoid any leakage from the top and the bottom of the cylinders, two steel plates (with rubber washers) were used. Figure 3. The cylinders filled with different volume of EG

5 3 rd International Conference on Concrete & Development / 603 When all samples were ready to be cast, as shown in Figure 5-33, the grout was poured in the 7 mm gap provided. The amount of the expansive grout (EG) that was used determined as a function of the amount of expansive agent (EA). Ratios (in weight) of 10%, 20%, 30% or 40% were used. Whilst filling the EG through the gap, vibrating by a smooth vibrator machine took place. The expansive grout was injected using a silicon gun with plastic pipe attachment with a diameter of five mm. 5. SPECIMEN DETAILS Since the chemical pre-tension (expansive pressure) is caused by the EA reacting against the confining jacket this means that the magnitude of this pre-tension depends on the degree of stiffness of the jacket and percentage of EA. Experiments were conducted by using these two parameters to quantify the amount of pretensioning on the jacket. Following that series of testing, concrete cylinders were confined with pre-tensioned composites. To vary the stiffness and strength of the confining jacket, one, two and three layers of carbon and glass sheets were used. Different percentages of EA were mixed with cement (10% and 20%) to achieve different confining pressures. Also, different materials for jacketing (Glass and Carbon) were used to achieve different stiffness and strength. A total of eighteen 100mm x 200mm concrete specimens were prepared without any pre-tensioning, 36 specimens were prepared with different levels of confinement pre-stressing and 15 unconfined specimen were tested under compression to determine the plain concrete strength. The concrete consisted of ASTM Type 1 Portland cement, river sand aggregate with a fineness modulus of 2.5 and gravel river aggregate with a maximum size of 10 mm. The water-cement ratio (w/c) was about 0.52 by mass. The average 28-days compressive strength of the concrete specimens was 31 MPa. Concrete specimens wrapped with one layer of Carbon and two layers of Glass without pre-tensioning force were designated as C0101, C0102, G0201 and G0202, respectively, whereas concrete cylinders with the same characteristics but pre-tensioned with 10% of EA and one layer of jacket were designated as C71101, C71102, G72101 and G72102, and with 20% of EA were designated as C71201, C71202, G72201 and G72202, respectively. Similar designation names were used for two and three layers either for non-pre-tensioned or pre-tensioned samples. 6. PREPARATION OF SPECIMENS For the wrapping of concrete cylinders without pre-tensioning, after applying the epoxy primer on the concrete surface, epoxy adhesive was applied and Carbon/Glass sheets were wrapped around the concrete cylinder until one wrapping layer was completed (with one third overlap). At the same time a special roller was used to help impregnate the fibre with resin and hardener and give a smooth finish. After curing, the strain gauges were glued directly onto the body of the jacket. For the pre-tensioned specimens, a gap is needed between the concrete and FRP for the insertion of the EA. For these experiments, the FRP jacket was premanufactured with a diameter 14mm larger than the concrete cylinder. After

6 604 / Application of Expansive Agent (EA) To Produce. curing, the jacket was placed around the concrete and the ends were capped to seal the expansive agent inside. Strain measurements were taken during the expansion phase of the expansive agent for up to four days. The testing procedure followed was then the same as for the unconfined specimen. 7. TEST PROCEDURE AND OBSERVATIONS The application of displacement to the specimens was controlled manually. The displacement was applied incrementally with each displacement level being held for a few seconds at each 0.1mm increment. All specimens were tested under centric (axial) loading. Failure was always explosive due to the high strain energy stored by the FRP material and it took place around the middle of the cylinder height as shown in Figure 4. Figure 4. Failure of the specimen confined with FRP jacket. 8. EXPERIMENTAL RESULTS The stress-strain diagrams for the specimen tested with one layer of either glass or carbon FRP are shown in Figure 5. Each graph shows the results for the unconfined cylinders as well as for the confined with or without pre-stressing. The right hand part of the graph shows the longitudinal strain whilst the left hand part shows the lateral strain. Strain measurements shown are the average values from the strain gauges and DV devices. The results are in general in good agreement between the two types of measurements, even though strain gauges measure local strains and DV devices integrate the strains over the length. The strain gauges also show the pre-strain that was developed by the EA. The longitudinal strain gauges on the pretensioned specimen show contradicting trend. Whilst the Carbon wrapped specimen show the strain to remain compressive, in the case of glass the strain eventually becomes tensile. This is partly because the glass wrapping contains longitudinal fibres as well, which restrain the concrete from expanding and lock some strain in the mid-height. It is also partly to do with the wrapping overlap,

7 3 rd International Conference on Concrete & Development / 605 which creates eccentric deformations on the specimen. Eccentric deformations and initial slip of the jacket may explain the tensile strains in the glass. 9. CONCLUSIONS This paper has shown experimental results from concrete specimen wrapped with glass and carbon FRP. In this investigation EA was used to produce post tensioning force in FRP jackets. It has shown that post tensioning can be used to enhance the load capacity and behaviour of concrete. REFERENCES 1. Mortazavi A., Strengthening, ductility and repair of RC structures by lateral confinement, MPhil thesis, University of Sheffield, Oct Weber & Broutin Ltd., En-Force strength in construction, SBD, 1999, UK. 3. Betonamit, The non-explosive cracking agent for universal application, 1998, UK, Re. No Pilakoutas K., Ductility design of reinforced concrete members, 16th European Seminar on Earthquake Engineering, Stara L., Czechoslovakia 6-12 Oct. 1991, pp Eurocode No 8, "Design provisions for earthquake resistance of structures", Part 1-4, "Strengthening and repair of buildings" (Draft), Commission of the European Communities, Brussels, 1995.