谁会翻译 钢结构期刊的

4. Tests on joints using extended endplates
The first series of joint tests was conducted on extended endplates which were
bolted to tubular column sections of various wall thicknesses. The comparisons were
conducted over a series of three tests in which the 8 mm, 10 mm and 12.5 mm
walled thickness were examined for the 200 3 200 serial sized SHS using grade
S275 material.
The first test in this group was joint test 21, where the extended endplate was
bolted to the 200 3 200 3 12.5 mm SHS. A 75 kN m moment was applied to the
joint before being unloaded. At this level of moment the joint responded linearly.
Fig. 3 shows the full experimental moment–rotation characteristic. The load was
steadily increased to give a moment of 236 kN m at which point the test had to be
stopped and restarted due to a problem with the hydraulic pump. When reloaded,
the joint stiffness was similar to the previously recorded unloading stiffness. The
moment-resistance of the joint continued to increase until the test was stopped at a
final recorded moment-resistance of 283 kN m (85% of the nominal fully plastic
moment of the beam) with a corresponding rotation of 0.042 radians. No bolt pullout
was observed for this test.
The joint exhibited excellent ductility, with rotations in excess of the maximum
rotations which would be found in practice. The moment–rotation characteristic is
clearly non-linear from an early stage of loading. Examination of the joint after the
test revealed that the majority of rotation of the joint had occurred due to flexural
action of the tube face with the endplate showing no signs of deformation. There
was, however, some yielding of the beam compression flange during the latter stages
of the test. The final deformed shape of the column after the test can be seen in Fig.
4, from which the location of the top row of the bolts in the extended endplate can
be observed. Extensive yielding of the tube face was found in this area which spread,
to a lesser degree, into the region around the second row of bolts. The yielding also
migrated into the webs of the column section by about 25 mm.
The bolt shown in Fig. 4 in the top row of the endplate was removed with great
difficulty, as the flowdrill thread had deformed appreciably and was close to thread
stripping before the test was stopped. The bolt was plastically deformed over the
lower length of the bolt thread as a result of the column face being distorted. The
stiffness of the bolt was not sufficient to restrain the column face when subjected
to a tensile load. The washer under the bolt head also showed signs of severe indentation.
A closer inspection of the hole revealed greater distortion to the top of the
flowdrill thread than to the bottom, resulting in a relatively small contact area
between the two threads. During the later stages of the test, the reduced contact area
lessened the resistance to bolt pull-out. It was also observed that the stiffness of the
heavy endplate kept the plate flat and the bolts perpendicular which induced a greater
stress on one side of the bolt thread as the tube wall deflected. Evidence of this can
be seen in Fig. 4 where the flowdrilled hole has distorted into an elliptical profile.
The remaining rows of bolts, closer to the compression zone of the joint, were
removed with relative ease.
The compression zone of the column (shown at the top of Fig. 4) was severely
indented around the profile of the endplate. Note the extent to which the endplate
indentation travelled, indicating the final position of the axis of rotation for this joint.
The webs of the column buckled outwards in this area with yielding being observed
across the full width of the webs. It is unclear if such extensive yielding would affect
the response of a joint connected to the adjacent face of the column as no tests were
conducted to investigate this 3D joint arrangement—or if such yielding would occur
if the endplates were bolted onto the column faces parallel to the beam.
The 283 kN m final moment capacity attained demonstrates that the joint qualifies
as partial strength, as the nominal moment capacity of the beam was 333 kN m. It
should be recognised that the extended endplate and beam combination is an extreme
test of moment capacity as the top serial weight of beam was adopted; with a lighter
beam the extended endplate and bolt group would be sufficient to mobilise the plastic
moment capacity of the section.
The second test of the series, test 19, featured an extended endplate connected to
a 200 3 200 3 8 mm SHS, the smallest wall thickness used in this series. A moment
of 35 kN m was first applied to the joint before it was unloaded and then reloaded
to failure. The full moment–rotation response is shown in Fig. 5. The sudden drop
in moment occurring at 60 milliradians was due to the top bolts stripping the flowdrill
thread and the bolts pulling out of the SHS. The maximum moment attained by the
joint before the bolts stripped was 162 kN m. After the bolts had failed, a lower
level of moment was resisted as the remaining bolts were still effective. At 75 milliradians
the test was stopped due to excessive deformation.
The moment–rotation curve for the joint exhibited a ductile failure for the practical
range of rotations before the bolts pulled out. The rotation of this joint was significantly
greater than that observed previously in test 21. It is probable that such a
failure would also have occurred for test 21 had sufficient rotation been applied.
The endplate was unbolted to facilitate examination of the column face. Some
difficulty was encountered in removing the bolts. During the unbolting process, the
already damaged threads of the bolt may have been distorted even further. The bolts
were not plastically deformed as in the previous test, indicating that the flexural
strength of the bolts overcame the rotation induced by the column face deformation,
unlike that of the 12.5 mm wall section. Nevertheless, the column faces had deformed
considerably more than those of the 12.5 mm thick column walled section. The
compression zone had also buckled outwards, whereas the tension zone of the joint
had extended to the top three rows of bolts, with equal severity of yielding observed
for the top two bolt rows. Although the endplate had not deformed, the extreme tips
of the beam compression flange had shown signs of yielding.
The third and final joint test in this series, test 20, was conducted with a 200 3
200 3 10 mm SHS column section. Thread stripping was again the cause of failure
but did not occur until the joint had reached a moment of 208 kN m at 56 milliradians.
Fig. 6 plots the moment–rotation envelope of test 20 together with the moment–
rotation responses of the two previous tests 21 and 19. The points indicated in Fig.
6 are selected to reflect the nature of the curve. The procedure of linearising the
curve is similar to that adopted previously for the series of simple joints [1] and
defines the outer envelope of the joints moment–rotation characteristic. Also shown
on this figure are indicative lines representing the linearised unloading stiffnesses
which are important in the context of column buckling in frames [3]. Examination
of the column specimen after the test revealed similar patterns of face deformations
to those of the previous two joint tests. Similarly, the extent of buckling in the
compression zone of the 10 mm walled column was intermediate between that exhibited
by the 8 mm and 12.5 mm column sections. As expected, the moment-resistance
of the flowdrill joint increased with the greater thickness of the column wall. The
initial stiffnesses of each joint were similar. The stiffnesses at 0.010 radians, just
after first yield, were surprisingly consistent, after which the joints exhibit a reduced
stiffness followed by a long plateau before reducing to virtually zero stiffness at the
end of each test. This characteristic of the moment–rotation curve indicated imminent
joint failure as the top row of flowdrill threads started to strip. In general, the
increased moment capacity was achieved at the expense of reduced ductility.