STRENGTHENING OF HEAT DAMAGED REINFORCED CONCRETE ELEMENTS

Abstract

In the light of extreme events of natural disasters (earthquakes or hurricanes) and accidents (fire or explosion), repairing and strengthening of existing concrete structures has become more common during the last decade due to the increasing knowledge and confidence in the use of advanced composite repairing materials. The past experience from real fires shows that it is exceptional for a concrete building to collapse as a result of fire and most fire-damaged concrete structures can be repaired economically rather than completely replacing or demolishing them. In this connection an experimental study was carried out to evaluate the performance of different repair schemes, which included the use of fiber reinforced polymer (FRP), ferrocement (FC), high strength fiber reinforced concrete (HSFRC) and steel plate (SP) jackets, in restoring the heat damaged structural concrete elements. A critical review of the literature shows that while tremendous amount of data is available on strengthening of reinforced concrete elements under normal ambient temperature condition, only limited studies have been carried out in the past on strengthening and restoring of heat damaged concrete elements. Previous researchers studied heat damaged reinforced concrete columns with one layer of wrapping of FRP and it has been reported that it does not provide much enhancement in the stiffness properties of strengthened reinforced concrete elements. Research on strengthening of heat damaged reinforced concrete elements with varying FRP parameters has not been attempted. Effectiveness of other available techniques of strengthening in restoring heat damaged concrete elements has not been investigated to best of knowledge. In the past strengthening of heat damaged RC elements has been studied by exposing the concrete elements to target temperatures of 500°C and 600°C only. However the influence of other temperatures of exposure has not been studied with respect to the behavior of strengthened heat damaged reinforced concrete elements. Earlier researchers have studied the effect of strengthening heat damaged RC elements, which were cooled under normal cooling conditions, but the rate of cooling in different cooling regimes like water sprinkling or water quenching was not studied. Influence of cyclic temperature conditions becomes important in some situations like cooling tower, cooling container, iii nuclear container vessels. But strengthening of such cyclic heat damaged elements has not been studied in the past. In view of the preceding discussion, the primary objective of the study undertaken in this research was to examine experimentally the effectiveness of various strengthening techniques in restoring structural performance of heat damaged RC elements. In this context, reinforced concrete columns and beams were constructed and then tested after grouping them into three main groups: un-heated, heat damaged and heat damaged & strengthened. The heat-damaged circular short columns, square short columns and T- Beams were initially damaged by heating to various temperatures and then were strengthened with different strengthening schemes. The variables of the study were type of strengthening scheme, heating temperature, number of heating cycles and type of cooling. In the first phase of the research program, a total of sixty-three reinforced concrete short columns were constructed and then tested after grouping them into three main groups: un-heated, heat damaged and heat damaged strengthened. All the specimens were of cylindrical shape (150 mm x 450 mm). The specimens were initially damaged by heating to various temperatures ranging from 3000C to 9000C. After exposing to target temperature for certain duration the specimens were cooled. The heating rate was set at 10°C /min., which was considered to be realistic for structures exposed to fire. Two different cooling regimes were chosen namely natural cooling and sudden water quenching. Some specimens were also subjected to cycle of heating and cooling. The heat damaged columns were subsequently repaired using different schemes namely FRP wraps, high strength fiber reinforced concrete jacketing and ferrocement jackets. The effects of various variables of the study were investigated with respect to stiffness, ductility, ultimate strain and the ultimate strength. The columns were tested under axial compressive loading. It was observed that the reduction in residual stiffness of heat damaged circular columns was greater than the reduction in ultimate load. The strength of heated specimens repaired with HSFRC jacket was higher than the heat damaged unstrengthen specimens, but remained less than the strength of control specimens in all cases. In case of FC jacketed specimens also, the axial compressive strength was higher than that of unstrengthen heat damaged specimens, but remained less than the control specimens except for 300°C heat damaged specimens. However, the strength of GFRP jacketed specimens was considerably higher than both the heat iv damaged unstrengthen specimen and the unheated control specimens. The stiffness of HSFRC and FC jacketed specimens was higher than both the control and heat damaged specimens exposed up to 600°C. However in the specimens heated to 900°C, though it was found to be higher than the heat damaged specimens but was less than that of control specimens. In case of GFRP, no significant improvement in stiffness of heat damaged specimens was noticed except in the specimen heated to 300°C. It was also observed that the specimens strengthened with GFRP, dissipated more energy compared to HSFRC and FC jacketed specimens. In the second phase of the research program a total of fifty-one reinforced concrete square columns were constructed and then tested after grouping them into three different groups: un-heated, heat damaged and heat damaged & strengthened. The specimens were initially heated to various temperatures of 300°C, 600°C & 900°C. After exposing to target temperatures for certain duration the specimens were cooled. The heat damaged columns were subsequently repaired with fiber reinforced polymer wraps, high strength fiber reinforced concrete jackets, ferrocement jacketing and steel plate jackets (SP). The effects of various variables of the study were investigated with respect to stiffness, ductility, ultimate strain and the ultimate strength. The square short columns were tested under axial compressive loading. The results show that the reduction in residual stiffness of heat damaged specimens was greater than the reduction in ultimate load. The results indicate that the strength values of GFRP jacketed specimens were higher than both the heat damaged unstrengthen specimens as well as the control unheated specimens at all the considered temperatures. However the specimens heated at 900°C temperature and strengthened with GFRP was not able to regain its original strength of the control specimens. The strength of heat damaged & HSFRC jacketed specimens was higher than 300°C, 600°C and 900°C heat damaged unstrengthen specimen up to 8%, 12% and 150% respectively, but remained less than the corresponding control unheated specimen at all the considered temperatures. Similarly, in case of FC jacketed specimens though the axial compressive strength was higher than the heat damaged unstrengthen specimen, but it remained less than that of the corresponding control specimens, except in 300°C heat damaged specimen. The restoration of stiffness of the heated damaged specimen depends upon the confining action of HSFRC, FC, GFRP and SP jackets. The confining action of GFRP jackets takes effect when the heated concrete specimen approaches its ultimate unconfined v compressive strength. The GFRP and SP do not contribute in the initial structural response because it is not active in the elastic range of the heated specimens when loaded axially. However, on the other hand, the HSFRC and FC jackets increased the stiffness of the heat damaged specimens due to the increase in the cross-sectional area of the specimens improving the dimensional stability of square columns. In the third phase of the research program a total of twenty-seven reinforced concrete T-Beams were constructed and then tested after grouping them into three main groups: un-heated, heat damaged and heat damaged strengthened. The beams were heated to temperatures of 600°C & 900°C. The beams were placed in the furnace upside down so that heat won’t affect the flange directly, which simulates the real condition of the structure during fire. The rate of heating was set at 10°C/min, which has been shown to be reasonable for structures exposed to fire. Each target temperature was maintained for three hours to achieve a thermal steady state condition. After heating it was observed that some hairline cracks developed at 600°C. The number of cracks became relatively pronounced at 900°C; therefore the section of beam had to be restored with micro concrete to make the surface a feasible one for strengthening. Subsequently the heat damaged beams were strengthened using the Glass Fibre Reinforced Polymer (GFRP), or High Strength Fibre Reinforced Concrete (HSFRC) and or Ferrocement (FC). The beams were tested in loading frame under 4 point loading condition using a 200 Ton capacity hand operated jacks connected to a data acquisition system through load cells. The beams were tested under monotonically increasing load. The deflection of the beams was noted using linear variable differential transducers (LVDT), placed at five locations at the bottom of beams and strain gauges were mounted on bottom of web and side web on GFRP jackets. The recorded data from the LVDTs, strain gauges and load cell were fed into a data acquisition system and stored on a computer. The load-deflection curves for the beams were examined to evaluate the effect of strengthening. The beams exposed to different temperatures caused a reduction in ultimate load carrying capacity by about 14 % to 61% when compared with undamaged control beams. The secant stiffness and energy dissipation were reduced by 34% to 56% and 10% to 41% respectively. The strengthened beams showed significant improvement in flexural strength over the heat damaged unstrengthen beams and in some cases the strength enhancement was even more than the undamaged control beams. The study proved that GFRP is the best option for the strengthening of heat damaged beams. On the contrary FC and HSFRC jacketing vi were mainly effective in improving ductility property but strength increase was minimal. The load corresponding to concrete cracking got increased considerably when the damaged beams were strengthened with different strengthening techniques. The fourth phase of research program includes an experimental study to investigate the bond behavior between GFRP laminate and heated concrete. The specimens were initially heated to various temperatures of 200°C, 400°C, 600°C & 800°C. After exposing to target temperatures for certain duration the specimens were cooled. The heat damaged specimens were subsequently bonded with fiber reinforced polymer sheets with various bond lengths. The test variables were different bond length; bond width and different elevated temperatures. The specimens were tested in single shear in a universal testing machine by dragging the GFRP sheet upward from the specimen and the applied load was measured with the help of a pressure sensor. The load was applied in a monotonically increasing manner. The bond strength increased with increasing bond length and reduced noticeably as temperature exceeded 400°C. The thickness of the delaminated concrete layer with GFRP composite was less for the specimens subjected to temperatures less than 400°C. However thickness of delamination was more in 600°C and 800°C heat damaged specimens. The value of the ultimate bond stress was influenced by the bond length and got increased with the decrease in bonded length. The failure of the specimens was found to be sudden followed by debonding of GFRP strip from the concrete. For specimen with 100 and 150mm bond length the increase in GFRP strain was gradual until 90 percent of failure load, but with larger bond length (200mm) the increase of GFRP strain was found to be gradual until 70 percent of the failure load. The propagation of debonding was clearly reflected by the strain distribution. A model has been proposed to estimate the bond strength between GFRP laminate and heated concrete.