Four knee joints were constructed and tested. The difference between t
he specimens was the connection detail. For three of the test specimen
s, beam and column longitudinal bars were mechanically attached to a s
tructural steel section embedded in the joint region. The fourth speci
men was a typical conventionally reinforced knee joint. The type of an
chors used for the test specimens would facilitate field installation.
All of the test specimens exhibit lower stiffness and load-carrying c
apacity when the joint is opened. This difference is more pronounced f
or the conventionally reinforced specimens. The response of the specim
ens under closing moments is similar for drifts up to 2.4 percent of i
nterstory height. The behavior is quite different under opening moment
s. At the interstory drift corresponding to 0.6 percent of story heigh
t, the load resisted by the conventionally reinforced specimen drops t
o 70 percent of that resisted by the specimens with composite details.
For composite specimens, a larger portion of the total energy was dis
sipated through beam flexural deformation. The failure of the conventi
onally reinforced specimen was due to excessive damage in the joint re
gion. The connections with composite details failed as a result of for
mation of ''plastic hinges'' in the beams, whereas excessive damage to
the connection region of the conventionally reinforced specimen lead
to failure.