Title:
Stress at Ultimate in Internally Unbonded Steel Based on Genetic Expression Programming
Author(s):
Nazar Oukaili and Iqbal Peera
Publication:
Structural Journal
Volume:
119
Issue:
6
Appears on pages(s):
177-191
Keywords:
flexural concrete member; genetic algorithm; post-tensioning; ultimate loading; unbonded steel tendon
DOI:
10.14359/51734803
Date:
11/1/2022
Abstract:
This study demonstrates the development of a new genetic
programming-based model to predict the increase in stress
Δfps beyond effective prestress, and the total value of stress fps
at ultimate loading in the internally unbonded steel tendons
of post-tensioned concrete flexural members. Genetic
expression programming (GEP) was employed to develop
an accurate predictive model for Δfps, summarizing almost
all the influencing parameters including the geometrical
properties of the structural member, the mechanical
properties of steel and concrete, and the loading type and
pattern into one single empirical expression. Comprehensive
experimental outcomes of 218 data points were compiled
from published literature since 1962 and the experimental
program carried out recently by authors on prestressed
concrete flexural members (beams and one-way slabs) with
internally unbonded steel. To inspect the potentiality of the
proposed model against 22 expressions suggested by various
researchers and Codes of Practice, an extensive comparative
study was carried out to assess Δfps. Compared to the other
predictive models, the proposed model showed the highest
correlation between the estimated and the experimental
stress increase Δfps in internally unbonded steel at ultimate
loading with a coefficient of determination R2 of 0.78.