Long slender reinforced concrete columns.
The behaviour of very slender eccentrically loaded
reinforced concrete columns is investigated, both
analytically and experimentally, under short-term and
Short-term loading investigations involved the testing
of thirty-six columns having slenderness ratios between
28.85 and 57.69. The first twenty-four columns were cast
with a concrete having a mean 28 day cube strength of 46 •. 82 N/mm2
and subjected to loading at eccentricities of 10mm and l5mm.
The remaining twelve columns were cast with a concrete having
a mean 28 day cube strength of 34.69 N/mm 2 and loaded at an
eccentricity of l5mm. The behaviour of four columns cast
with the stronger of the two concretes was monitored under
sustained loading for a period of about eighteen months.
These columns had slenderness ratios of 38.46 and 48.08 and
loads were applied at eccentricities of 10mm and l5mm.
An analytical procedure based on the classical equations
of mechanics (equilibrium compatibility and properties of
the constituents) has been developed to predict the behaviour
of slender columns under an eccentric load.
Comparisons between the observed and analytical
behaviour of forty columns, indicate that the proposed
analysis estimates closely the load capacity and load
deformation behaviour of very slender reinforced concrete
columns. The validity of the proposed method has also been
demonstrated in extensive comparisons between predicted
and experimental observations obtained from tests performed
by other investigators on very slender columns.
The design of very slender columns in accordance with
CPllO is based on the assUmption that failure would occur
when the maximum concrete compressive strain reaches an
ultimate value of 0.0035. In tests this assumption has
proven to be wrong as instability occurs at very low concrete
strains, normally no more than 0.001. Thus the code
mentiqned is incorrect in using a theory based on material
failure· for very slender columns.
The results of analytical studies into the behaviour
of slender columns under sustained loading indicate that
the present provisions in CPllO do not account fully for
creep effects. As a consequence the recommendations of
CPllO lead to unconservative designs. The recommendations
of ACI318 on the other hand, agree in general with the