22-10-2012, 01:54 PM
SEISMIC RETROFITTING OF EARTHQUAKE-DAMAGED CONCRETE COLUMNS BY LATERAL PRE-TENSIONING OF FRP BELTS
SEISMIC RETROFITTING.pdf (Size: 649.73 KB / Downloads: 121)
ABSTRACT
Five square columns with two shear span-to-depth ratios of 1.5 and 2.5 were
constructed to model half-scale shear-deficient columns and tested under constant
axial compression and reversed cyclic lateral load, simultaneously. After being
tested, two of the columns with different shear span-to-depth ratios were
subjected to a certain level of damage in terms of crack pattern and also drop in
the lateral capacity. Then, these earthquake-damaged columns were retrofitted by
pre-tensioned carbon or aramid FRP belts, and once more, were tested under
cyclic lateral loading and constant axial compression. As the confining devices,
i.e. FRP belts, were pre-tensioned before applying the lateral load to the columns,
both active and passive confinements were utilized. As an instant result of pretensioning,
the initial cracks of the damaged column were closed. It should be
noted that this retrofitting procedure is quick as it is carried out without any
repair measures such as removal of damaged concrete or crack injection and so
on. Moreover, the prestressing technique is an innovative method and can be
applied manually using a simple wrench. According to test results, the lateral
capacity of the original columns dropped suddenly, showing a brittle shear
failure. When the damaged columns were retrofitted by pre-tensioned FRP belts,
the lateral strength could be restored and the drop in shear capacity could be
prevented up to large drifts, indicating a better seismic performance.
Introduction
In order to reduce earthquake disaster, three countermeasures are recognized, as follows:
Mitigation/Prevention, Preparedness/Emergency Response, and Recovery/Reconstruction Plan.
Recent damaging earthquakes such as Northridge (1994) and Hanshin-Awaji (1995) have
clearly revealed that the highest priority for seismic hazard reduction is mitigation and structural
issues. In other words, various problems generated after the quake might not have become so
severe if structural damages were much less.
While retrofitting is most likely referred to the non-damaged existing buildings, there is
also a need to upgrade the structures which are partially damaged following an earthquake. It
goes without saying that rehabilitation of earthquake-damaged buildings can be a case only if
the level of damage is not that high. In this capacity, the objective of rehabilitation can be either
a permanent solution for enhancing the seismic performance of damaged buildings up to the
desirable level of performance or as a temporary solution to avoid further damages to the
earthquake-affected structures during the possible aftershocks or even upcoming quakes.
Details of Retrofitting Technique
An innovative technique is proposed in order that FRP belts are prestressed manually
using a simple wrench (see Fig. 1). The technique is as follows: a carbon or aramid fiber belt is
cut in a desirable length needed for wrapping around the column cross section, and then, is
impregnated with epoxy resin along only 100 mm lap joint of both cut ends to form a loop,
which is straightened to form a two-ply belt. Each end of the straightened two-ply belt looks like
an eye-hook, through which a steel rod (with threaded holes at its both ends) can be passed.
When the two-ply belt is wound around the column, its both ends can be clamped together by
putting a couple of rods into the end eyes of the belt, and then, passing bolts through the rod
holes. Then, prestressing can be given to the belts by manually screw driving the bolts. Herein,
prestressing is applied before the belts are impregnated with the epoxy resin (except for the 100-
mm lap joint), and thereby, dry fibers are wrapped around the square section. Although dry
fibers are used during prestressing procedure, it is recommended to impregnate the belts (after
pre-tensioning) for long-term applications.
Outline of Experimental Program
Table 1 shows the test parameters for five square columns with the dimension of 250 mm
and two shear span-to-depth ratios of 1.5 and 2.5. The axial force ratio (i.e. the axial stress
applying on the gross sectional area of the column divided by the concrete cylindrical strength)
is equal to 0.2 for all specimens. The geometrical scale factor for these columns is about 2.4 to
model the low-rise concrete buildings which were designed in accordance with old seismic
codes and were basically shear deficient columns because of poor arrangement of internal steel
ties. The column with shear span-to-depth ratio of 2.5 was retrofitted by use of three carbon
fiber belts distributed evenly along height of 1.5D next to the base of the cantilever column
(where D is the dimension of cross section, i.e. 250 mm). The columns with shorter shear spanto-
depth ratio of 1.5 were retrofitted using the pre-tensioned aramid fiber belts, which were
evenly distributed along total height of the columns and were located on steel angles at the
corners of cross section (as indicated in Fig. 1).
Experimental Results
The non-retrofitted test column RM1-0 was tested under the drift ratios of R= 0.5, 1.0,
1.5 % (with three cycles for each), and 2.0, 2.5, 3.0 % (with one cycle for each). Then the lateral
displacement increased up to R= 4.0% when the shear capacity suddenly dropped, and at about
R= 4.5% the drop in lateral strength of the column became more than half of the peak strength
(see Fig. 2). Moreover, diagonal cracks appeared in the column face, showing a brittle shear
failure. Hereafter, this earthquake-damaged column was about to be rehabilitated by pretensioned
carbon fiber belts and re-tested under constant axial compression and reversed cyclic
lateral displacements. But before re-testing, the corner concretes, which were peeled off during
the first stage of the cyclic loading test, were repaired by using cement mortar so as to provide a
smooth surface for placing the belts. This repair operation was very simple and quick as it was
not accompanied with replacement of damaged concrete and also was not involved with crack
injection and so on. In the next stage, only three carbon fiber belts with the intervals of 75 mm
were wrapped around the column and distributed next to the base of the cantilever column. The
lateral pre-tensioning strain equal to 2500 microns (i.e. a sixth of carbon fiber ultimate strain)
was applied to the carbon fiber belts. As an instant result of the lateral pre-tensioning, the
cracks, which remained open in the damaged column, were getting closed. Hereinafter, the
retrofitted column was re-named as ERM1-C75/6 and was ready to be tested.
Conclusions
This paper presented experimental results for retrofitting of earthquake-damaged square
concrete columns by use of lateral pre-tensioning of FRP belts. As in the proposed procedure
neither the damaged concrete was replaced by new concrete nor any repair operation such as
crack injection was employed, the technique was quick in application and was described as an
emergency retrofit. One of the test columns with shear span-to-depth ratio of 2.5 was tested
under constant axial compression and reversed cyclic lateral displacements up to drift ratio of
4% where its lateral capacity suddenly dropped. Thereafter, the damaged part of the column (i.e.
the part next to the column base) was wrapped by a few prestressed carbon fiber belts. Another
test column with shear span-to-depth ratio of 1.5 which was partially retrofitted was tested up to
drift ratio of 2.5% where the lateral capacity dropped by 20% and some damages in the form of
cracks appeared, and then once again, was retrofitted by additional prestressed aramid fiber
belts.