Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE ELEVENTH INTERNATIONAL CONFERENCE ON CIVIL, STRUCTURAL AND ENVIRONMENTAL ENGINEERING COMPUTING
Edited by: B.H.V. Topping
Statistical Modeling of Post-Heating Residual Strength for Portland Cement Concrete
R.H. Haddad and A. Qudah
Department of Civil Engineering, Jordan University of Science and Technology, Irbid, Jordan
R.H. Haddad, A. Qudah, "Statistical Modeling of Post-Heating Residual Strength for Portland Cement Concrete", in B.H.V. Topping, (Editor), "Proceedings of the Eleventh International Conference on Civil, Structural and Environmental Engineering Computing", Civil-Comp Press, Stirlingshire, UK, Paper 162, 2007. doi:10.4203/ccp.86.162
Keywords: concrete, fire, heat, prediction, strength.
Strength estimation of concrete in fire-damaged structure enables engineers to propose appropriate repair strategies, if repair is practical and economically visible, or recommend demolishing. In the field, strength is estimated either using concrete cores (obtained from heat-damaged structures) or by employing certain non-destructive techniques. Concrete coring is destructive and expensive whereas current non-destructive techniques are limited in their ability to detect damage and hence estimate concrete strength accurately. In this paper, a non-linear statistical model is proposed to predict the residual compressive strength for Portland cement concrete (without pozzolanic additives). The parameters of the model include exposure temperature, water-cement ratio, type and weight fraction of aggregates, relative humidity, and testing conditions.
The model developed is based on literature data which were mainly concerned with mechanical behavior of ordinary and high strength concretes under high temperatures in the range (100-1000oC). The data used covers water-cement ratios of (0.25-0.7), weight fraction for two types of aggregate (siliceous: basalt and granite, and carbonic: limestone), relative humidity (50-100%), two testing conditions (testing while hot, and after cooling) and different volumes of test specimens.
The predictability of the model may be ranked as very good. The multiple coefficients of determination (R2) and the proportion of variance accounted for were found to be 0.96 and 0.93, respectively. Moreover, the trend behaviors of the model's residual strength versus temperature are in agreement with those obtained experimentally and with Eurocode and CBE designing curves.
The sensitivity of the residual compressive strength to the model's parameters can be summarized as follows:
purchase the full-text of this paper (price £20)