IMPROVING DUCTILITY OF NON-SEISMICALLY DESIGNED RC COLUMNS

IMPROVING DUCTILITY OF NON-SEISMICALLY DESIGNED RC COLUMNS : 

IMPROVING DUCTILITY OF NON-SEISMICALLY DESIGNED RC COLUMNS

Slide 2: 

An experimental investigation of the improvement of ductile behavior of non- seismically designed reinforced concrete (RC) columns with various configurations of transverse reinforcement is presented in this paper. Emphasis is placed on the effects of the configuration of transverse reinforcement on the enhancement of ductility and improvements in seismic performance of RC columns. It has been shown that adopting the combination of non-seismic detailing of links with the proposed transverse ties in the column sections can effectively enhance the ductility capacity and energy dissipation ability, thus improving the seismic behavior of non-seismically designed RC columns, while the construction cost and difficulty will not be increased significantly.

INTRODUCTION : 

INTRODUCTION The previous earthquakes indicated that although the seismic intensity of the earthquakes is not high, it could still cause a great loss of life and economic damage in a region of moderate seismicity, where no urban earthquake disaster management program is in place. In general, the peak ground acceleration (PGA) in regions of low to moderate seismicity is recognized to be about 0.1 to 0.15 g with a 10% probability of exceedance within 50 years. Earthquake threat is normally not a major consideration for building and structure design in these regions; while the structures are, in general, only designed non-seismically to carry their service loads, that is dead, imposed and wind loads.

Slide 4: 

The seismic design for severe earthquake regions requires special steel detailing for reinforced concrete (RC) structures in order for the structures to behave in a ductile manner and dissipate the earthquake energy by inelastic action. For RC columns, high reinforcement ratio and close spacing of stirrups are generally required to provide adequate confinement for the concrete. For a moderate level of seismic design, it would not be wise to adopt directly the same design as for a strong earthquake risk, which may lead to a tremendous increase in construction cost, time and difficulty.

Slide 5: 

Research has indicated that the strength & ductility of RC columns increase with core concrete confinement and are influenced significantly by the tie configurations. This paper reports tests on non-seismically designed RC columns with various configurations of transverse reinforcement with horizontal ties. The study aims at investigating the ductility capacity of RC columns with non-seismic reinforcement detailing combining the proposed additional configurations of transverse reinforcement.

Slide 6: 

It has been shown that using the proposed configurations of transverse reinforcement, but maintaining a similar stirrup ratio, can effectively enhance the ductility capacity and energy dissipation ability, thus improving the seismic behavior of non-seismically designed RC columns.

2. EXPERIMENTAL PROGRAMME : 

2. EXPERIMENTAL PROGRAMME In the experimental program, six RC column specimens are tested under simulated seismic loading and a constant axial compression. The variables investigated include the configuration, spacing and ratio of transverse reinforcement. The cross sections of specimens are illustrated in Fig. 1. A summary of test specimens and their properties is presented in Table 1.

2.1 TEST SPECIMENS : 

2.1 TEST SPECIMENS All the specimens had square cross sections of 300x 300 mm and a total height of 1.5 m. A basement block of 500 x 1000x 1000 mm, which simulates an adjoining footing or cap, was cast at the base of the column. The column specimens were designed with different transverse reinforcement ratios and configurations. Details of the geometry and reinforcement detailing are shown in Fig. 2. According to the non-seismic reinforcement detailing requirement, the spacing of links in a RC column should be equal to or less than 300 mm, 12 times the bar diameter, or 0.75 times the depth of the section, whichever is the smallest.

Figure 2. Geometry and steel details of specimens: (a) specimens 1 and 2; (b) specimens 3 and 4; : 

Figure 2. Geometry and steel details of specimens: (a) specimens 1 and 2; (b) specimens 3 and 4;

Figure 2. Geometry and steel details of specimens (c) specimen 5 (d) specimen 6 : 

Figure 2. Geometry and steel details of specimens (c) specimen 5 (d) specimen 6

Slide 13: 

For the test specimens, 0.75 times the depth of the section (225 mm) became the control. Link spacing of 200 mm was adopted for specimen 1 (NSD-200). Specimen 2 (NSD-130) was designed with closer link spacing (130 mm) along the height of the column in order to achieve better ductility. Specimen 3 (MNSD-200) adopted the proposed modified non-seismic detailing, which involves insertion of additional transverse ties into the cross section.

Slide 14: 

The additional transverse ties are anchored by 900 and 1350bends around the bars, plus an extension of eight tie bar diameters at each side respectively. They are arranged alternatively through the height of the column. Here the link spacing is 200 mm. However, the transverse reinforcement ratio is about the same as that of specimen 2.

Slide 15: 

Details of the transverse steel ties of specimen 4 are the same as those of specimen 3, but with closer link spacing of 130 mm assigned along the full height of the column so that a comparison can be made with specimens 2 and 3 for the confinement effect achieved through the use of additional transverse ties and closer link spacing respectively. In seismic design, the links are required to be anchored by a 1350 bend with at least 8 times the bar diameter extension. These links have to be provided through the height of columns and even closer at the potential plastic hinge region.

Slide 16: 

Both specimen 5 (NSD-200-PH-50) and specimen 6 (MNSD-200-PH-50) take the POTENTIAL PLASTIC HINGEs into consideration. Link spacing within the plastic hinge zone is reduced to 50 mm with insertion of the ties. In the remaining part of the column, there are no transverse ties in specimen 5, while specimen 6 keeps those ties in the cross section. The perimeter hoops in all specimens are 900 ones anchored with 8 bar diameter extension.

2.2. Test setup and procedure : 

2.2. Test setup and procedure The test setup and loading system are shown in Fig. 3. The specimen was mounted on the strong floor of the laboratory. Axial load was applied to the specimen through the vertical actuators that are connected at a pair of loading frames, which can be moved together with the specimen during the experiment. Lateral load reversal is then applied by a 250 kN servo DARTEC actuator, which is supported by the strong wall in the laboratory.

Figure 3. Test-rig and test setup : 

Figure 3. Test-rig and test setup

Slide 19: 

The stroke of the actuator is 200 mm, 100 mm each in positive and negative directions respectively. In the experiment, both load control and deflection control modes were used at different stages. Under the load control process, the specimen undergoes increasing cycles of horizontal load until the rate of change in displacement is very high at a relatively constant load.

Slide 20: 

It is usually difficult to observe such a phenomenon during the experiment, and in general the control method is changed when 80% of the predicted ultimate strength of the specimen is reached. The subsequent displacements applied are dependent on the previous yield displacement and were increased gradually until failure of the column. It is assumed that failure occurs when the specimen’s ratio of the restoring force at the maximum displacement to that at yield is around 0.8.

Slide 21: 

However, all the column specimens were tested until extensive cracking and spalling occurred. Three 100 s cycles of loading or displacement are adopted. The hysteretic loops of applied lateral loads versus the column free end displacement were continuously plotted and updated in the computer during the test.

2.3. Instrumentation : 

2.3. Instrumentation The deflection of the tip of the column is captured through Linear Variable Differential Transformers (LVDTs). They were put on a steel platform standing on the basement block of the column. In order to determine the strain in reinforcement steel, electrical-resistance strain gauges were attached at different locations on the longitudinal and transverse reinforcement steel.

Slide 23: 

In addition, two pairs of LVDTs with a range of 25 or 50 mm were placed on each side of the column close to the base, to measure the vertical deflections . Thus we obtain the distributions of curvature and rotation in the potential plastic hinge region.

3. RESULTS3.1. General observation : 

3. RESULTS3.1. General observation All specimens failed in flexure. The horizontal load recorded at first appearance of flexural cracks varied from 60 to 80 kN and the ultimate load varied from 108 to 128 kN. In the final stage of the experiment, all specimens exhibited different behaviours. Main diagonal cracks formed in the body of specimen 1 (NSD- 200). Surface concrete became crushed and spalled off with the corresponding load decrease.

Slide 25: 

For specimen 2 (NSD-130), the 90o anchorage of links opened and buckling of the reinforcement was observed after spalling of surface concrete. Specimens 3–6 all exhibited stable strength and stiffness degradations during the experiment. The extent of cracking was small when compared to those of the first two specimens. The cracks mainly concentrated on the potential plastic hinge region (Fig. 4).

Figure 4. Failure of specimens: (a) NSD-200 (b) MNSD-200 : 

Figure 4. Failure of specimens: (a) NSD-200 (b) MNSD-200

Figure 4. Failure of specimens (c) NSD-200-PH-50 (d) MNSD-200-PH-50 : 

Figure 4. Failure of specimens (c) NSD-200-PH-50 (d) MNSD-200-PH-50

3.2. Hysteretic behaviour : 

3.2. Hysteretic behaviour The horizontal load versus horizontal displacement relationship of the specimens is illustrated in Fig. 5. From the hysteretic loops the strength and stiffness degradation can be observed. The energy dissipation capacity can also be accounted for. For comparison, all the graphs are plotted with a maximum displacement of 100 mm in both positive and negative directions.

Slide 30: 

It was observed that the area of hysteretic loops in specimen 1 was small and thin. The strength and stiffness reduced rapidly after certain cycles of loading. The hysteretic loop areas of specimens 2, 3 and 4 increased, the strength and stiffness degradation was slow; while the hysteretic responses of specimens 5 and 6 showed that the specimens have a stable behaviour, good energy dissipation and a limited reduction in strength up to the final stage of the test.

3.3. Ductility factor : 

3.3. Ductility factor The displacement ductility factor can conveniently be defined as a ratio of displacement at a point corresponding to 80% of the maximum horizontal load plotted against displacement at the occurrence of longitudinal reinforcement bar yielding. Owing to difficulties in determining the overall yielding point of a cross section, the first yield displacement is generally used as the reference for the ductility factor determination. .

Slide 32: 

In addition to displacement ductility, it is common to evaluate ductility capacity of structural elements by a curvature ductility factor. In the same fashion, the first yielding curvature of the section for both directions is determined. Table 2 presents the experimental results of the yielding and ultimate displacements in the horizontal direction. Table 3 shows the corresponding yielding and ultimate curvatures.

3.4. Energy dissipation capacity : 

3.4. Energy dissipation capacity The area inside the hysteretic loops can be considered as an indirect measure of the energy that is dissipated by the plastic hinge. It is obvious that the larger the area occupied (that is, the higher the deformation level of the element), the larger the dissipated energy and the larger the damping effect provided. Values for energy dissipation are obtained by measuring the area inside the hysteretic loops at an ultimate load of 80% of the maximum load.

Slide 35: 

Normalised energy is also considered for comparison. It is defined as the energy obtained divided by that of specimen 1. The energy dissipation capacities of the specimens are given in Table 4.

4. DISCUSSION : 

4. DISCUSSION In the experiment, specimen 1 was designed as a control specimen, to indicate the behaviour of an RC column with typical non-seismic reinforcement detailing. It was observed that the seismic performance of specimen 1 was the worst of all the columns. From a ductility analysis, although it shows ductile failure mode, the ductility factor value is minimal.

Slide 38: 

From an energy analysis, it achieves just 47.6% of specimen 3 and only 25.9 and 19.4% of specimens 5 and 6, respectively. The earthquake resistance ability of specimen 1 is limited and no obvious plastic hinge could be observed. Instead, wide diagonal cracks are formed at the final stage of testing.

Slide 39: 

In order to compare the effectiveness of links and transverse ties, specimen 2 keeps the same transverse reinforcement ratio as that of specimen 3, but the spacing of links was decreased from 200 to 130 mm. Both displacement and curvature ductility factors obtained indicate that specimen 3 has a slight advantage over specimen 2. From an energy analysis, both specimens perform better than specimen 1.

Slide 40: 

However, specimen 3 dissipated more energy than specimen 2. Also, at the final stage of specimen 2 testing, the 90o anchorage of the links opened with resultant buckling of the reinforcement and the strength was reduced drastically after spalling and crushing of the concrete. In contrast, a stable reduction in strength was observed in specimen 3 and reinforcement buckling was not obvious during the experiment.

Slide 41: 

By adding two transverse ties on each of the reinforcement cross section and keeping the same link spacing of a non-seismic design (200 mm), the seismic performance of columns was improved. The measured displacement ductility factor increased by around 90% and the measured curvature ductility improved by more than three times when compared to specimen 1. The energy dissipation ability dramatically increased by more than 100%.

Slide 42: 

In addition, it was observed that a plastic hinge forms at the lower quarter of the column. Specimen 4 adopted the modified non-seismic detailing with spacing of 130 mm. It consisted of closer spacing compared with specimen 3. It exhibited better ductility and energydissipation. The increment of transverse steel reinforcement ratio was about 53%.

Slide 43: 

Ductility increased from 3.12 to 4.06, an improvement 30%, while energy dissipation improved by 68%. On the other hand, specimen 4 included additional transverse ties compared specimen 2. However, specimen 4 exhibited a 45% increase in ductility and a 100% increase energy dissipation.

Slide 44: 

With the adoption of the plastic hinge reinforcement concept in a strong earthquake region, the links in specimen 5 were arranged as close as 50 mm in the lower section (450 mm) of the column. The transverse ties were still provided for a better confinement effect. This specimen performed very well with stable energy dissipation and reduced stiffness degradation. Specimen 5 was reinforced in the same way as specimen 3, but there were no transverse ties in the area above the plastic hinge region.

Slide 45: 

When considering the measured displacement ductility and curvature ductility factor, specimen 5 out performs both specimen 1 and specimen 3, while specimen 3 behaves similarly to specimen 5. Under an energy dissipation analysis, specimen 5 dissipated more energy than specimens 1, 2 and 3. However, it managed around 75% of the total energy dissipated by specimen 6.

Slide 46: 

Out of all six columns, specimen 6 performs excellently in all aspects. The ductility factor was increased to 5.10, which is more than threefold that of specimen 1 and the curvature ductility factor was more than fourfold that of specimen 1. From an energy dissipation analysis, specimen 6 dissipated 17.52 kNm, more than five times that of specimen 1. The reduction in strength was stable and limited damage was observed during the experiment.

Slide 47: 

It has also been shown from the results that closer spacing of links provides for higher ductility levels. This is particularly important when strong earthquake resistance is required. For the same transverse steel ratio, configuration of the reinforcement becomes crucial to the overall seismic performance of the column. It can be seen from the results that adding transverse ties is more effective than closing of thelinks when ductility improvement is sought. The authors believe that both transverse ties and additional links support longitudinal reinforcement between corner bars and confine the concrete. Overall confinement is essential for ductility enhancement of concrete columns.

5. CONCLUSION : 

5. CONCLUSION (a) Column specimens with non-seismic reinforcement detailing possess low ductility capacity and poor energy dissipation. Rapid strength and stiffness degradation was observed during the experiment. (b) By maintaining the same transverse reinforcement ratio, the seismic performance of RC columns with additional transverse ties is superior to that with closer spacing of links in which buckling of longitudinal bars and opening of transverse steel anchorage was observed. The ductility of the column specimens can be enhanced significantly by the addition of transverse ties.

Slide 49: 

(c) With consideration of the potential plastic-hinge region, the energy dissipation and ductility capacity was improved tremendously. Stable degradation of strength and stiffness, and limited cracking and crushing of cover concrete was recorded compared to other specimens. (d) Ties with 900and 1350 ends perform very well in collaboration with links arranged according to general non-seismic design criteria. They can also effectively restrain possible buckling of unsupported longitudinal reinforcement bars, which indirectly confines the core of the column.

5.1. Design recommendations : 

5.1. Design recommendations In order to achieve confidence in the earthquake resistance of RC columns in a moderate seismicity region, the adoption of extra 90–1350 transverse ties or extra links, such as specimen 1, offers an effective way to enhance the ductility and energy dissipation capacities. It has been shown that by adding additional transverse ties a doubling of both energy dissipation capacity and displacement ductility, in addition to a tripling of the curvature ductility can be achieved, while the increase in construction cost is less than 4%.

Slide 51: 

Based on the non-seismic reinforcement detailing requirement plus the additional transverse ties, the ductility enhancement would be enough to satisfy the requirements of low to moderately seismic regions. The cost and difficulty of construction would not be increased significantly by using the proposed detailing technique. For strategically important buildings, such as hospitals and government offices requiring higher levels of safety, a higher ductility capacity may be required.

Slide 52: 

By implementing the reinforcement detailing of specimen 5, a fourfold increase energy dissipation can be achieved compared to columns with typical non-seismic detailing. I If better seismic performance than that exhibited by specimen 5 is required, transverse ties can be added throughout the entire length of the column, including the plastic hinge region and remaining portion, such as those installed in specimen 6. With such detailing a fivefold increase in energy dissipation and more than threefold increase in ductility capacity is achievable.