OPTIMIZATION OF PRESTRESSED WAFFLE SLABS VIA GENETIC ALGORITHM

Resumo

In the majority of current buildings, the use of prestressing has enabled the design of projects
with larger spans than the conventional ones. Besides, the technique has favoured greater durability of
the structures, since it avoids problems of cracking and excessive deformation in service. However, most
projects are still heavily guided by the expertise of the structural designer, which does not guarantee
optimal projects from an economic point of view. This work aims to present an optimization formulation
for waffle slabs executed with current plastic moulds, using the heuristic of genetic algorithms in the
program ‘BIOS‘ (Biologically Inspired Optimization System). The analysis of the efforts was integrated
into the model using the ‘FAST‘ program (Finite Element Analysis Tool). To evaluate optimum changes
in the geometric characteristics of plastic moulds, currently provided in commercial catalogues, design

variables related to the cross-section geometry of the slab, the number of tendons per rib and their ec-
centricity in the middle of the span were selected. Thus, the application of the evolutionary algorithm

seeks to minimize the objective function of the model, composed of the costs with the materials and
labour used, while attending to the constraints of ultimate and service states in the design, arranged in
standards. In all cases, the slabs are simply supported and the prestressing was considered as equivalent
load. A comparison is made between the costs of the formulated optimization model and another one that
incorporates the geometry of the plastic moulds currently available in the market. The outcomes indicate
that reasonable reductions can be achieved in the total execution costs of the waffle slabs by varying the
input parameters of the algorithm, such as the spans adopted. Also, is evaluated the effect of adopting
slabs with higher spans ratio over the optimization results.