Prediction of Axial Load Capacity of Reinforced Concrete Column Using Artificial Neural Network

The axial load capacity (ALC) of a reinforced concrete (RC) column is a fundamental aspect of a structure. Thus, the systematic computation process plays a vital role in building design. It is widely encouraged to use artificial neural network (ANN) for this uncommon problem to predict the precise outcome. This study used three different algorithms of the ANN model to forecast the ALC of an RC column provided by its specific properties. These are Bayesian regularization backpropagation (BRB), Levenberg-Marquardt, and scaled conjugate gradient. The given dataset from the literature was trained in MATLAB software, and it was found that the BRB algorithm has a more positive performance compared to other algorithms. This study presents a detailed methodology and results of the model. Utilizing ANN would provide a more accurate and robust tool for predicting and giving new recommendations for professionals to design safer, more economical, and more efficient structures.

I. Introduction

Due to its durability and accessibility, most buildings and structures in the Philippines are made up of reinforced concrete (RC). This is due to the ease and affordability of production and the universal availability of the primary ingredients [1] . RC is widely used for minor or major construction. Due to its strength with the support of steel bars embedded inside the concrete, structures can withstand greater loads, which is why it is also used for creating columns. With the proper concrete admixtures and mixed design, today’s concrete is simple to handle on the job site and may satisfy a variety of specifications of the construction project. Like any construction building, the columns of an RC structure are substantially correlated to the axial load (AL). An AL is a force that acts along a structure’s axis or focal point and generates a consistent effort [2] . When a structure experiences an AL as a result of compression, a force acting in the same direction tends to crush the structure. To maintain the stability of the structure, this calculation is essential to determine how much of this can withstand and its capacity. Reducing the axial force in columns optimizes their sizes and reduces their self-weight, making them easier to utilize in buildings [3] .