الفهرس 
ONCRETE COMPRESSIVE STRENGTH (NSC & HSCOnly 14 pages are availabe for public view
 |  | from | 32 |  |  |
| | | from | 32 | | |
Abstract
High performance Concrete (HPC) is known as a high technology construction material, proving to be very cost effective, reliable, and having long term durability in natural environment. Designers are termed upon to design slimmer, lighter structures with high load carrying capacities. These slimmer structures are much more disposed to punching shear failures, especially in the area of column/slab connections.
Reinforced concrete plates exhibit complexities – most importantly punching shear - in their structural behaviour. This is due to the composite nature of the materials and factors affecting such behavior. Unanimous agreement states that the factors of most significant effect – for the overall punching shear capacity - are namely concrete compressive strength fcu, flexural reinforcement ratio ρ and its yielding strength fy, effective slab depth d, and the effective perimeter bo (a function of the column geometry and slab depth). Concrete compressive strength, however, is the most prominent of the above. Researchers have emphasized on the usage of high performance concrete HPC– particularly high strength concrete HSC – to observe its effect on the punching shear capacity of the slab-column connections.
In this regards, a thorough literature review was conducted on the punching shear phenomenon in slab-column connections and influencing factors. A total of 495 punching shear sample results were collected from a diversity of research studies. In this study, the data-set is used to compare the already existing major formulas in international codes as well as research efforts; predominantly; based on empirical equations. The coefficients of these models are thus re-calibrated (optimized using the Matlab optimization toolbox); giving – in certain cases – a significant enhancement to the model performance. A new empirical (optimization-based) equation is proposed, yielding excellent performance (minimum sum square error); considering the 495 samples.
Furthermore, an Artificial Neural Network (ANN) model is developed using the collected data. The results obtained show high proximity to experimental output punching shear capacity results. Results on the ANN model are finally compared to the punching shear capacity of the proposed model that gave the best results after coefficient recalibration. This thesis also includes a parametric study for the punching shear input parameters and their significance as regards the output punching shear capacity of the slab-column interior connections.