Abstract:
Conventional modeling of reinforced concrete structures by FE method
results in too long and unrealistic value of period of buildings. A reason
of this phenomenon is that a structure usually contains many secondary
elements likeI infills, which are generally ignored to have any structural
contribution. lAs such the numerical models become too flexible resulting
in longer time period. For this reason, code provisions usually impose a
limit on the period value obtained from numerical analysis with
conventional FE modeling. In this thesis, an extensive computational
study has been conducted to determine the fundamental period of
buildings considering the structural effect of infills.
A numerical investigation has been performed to calculate the fundamental
period of a series of regular building frames under various conditions.
Variation of number of floors, floor height, number of span, number of bays,
floor panel size, percentage of infills etc. are considered during the
investigation. The infills are modeled as equivalent diagonal strut. Beams
and columns are modeled using three-dimensional frame element and floor
slab is modeled using shell element. The period of the same buildings are
also evaluated using empirical code equations and a comparison with the
results of FE analysis are made.
On the basis of the investigation it has been found that the period found by
modal analysis for models without infill are much higher than the same
found by code equations and from modal analysis with infill. It has been
found that the code equations generally produce shorter values of periods
when compared to the periods obtained from modal analysis with infill,
although the difference is not as high as with the models without infill.
Three parameters namely (i) number of stories or total height of the
building, (ii)floor panel size and (iii)amount of infilled panels present in the
building have been identified to have significant influence the building
period. Incorporating the effects of these parameters, a few correction
factors have been proposed which may be used to further refine the period
value predicted by code equations.
The period of a few arbitrary examples are determined on the basis of the
proposal and are compared with modal analysis. It has been found that the
periods determined by applying the suggested correction factors match
better with modal analysis than existing code equations.
