Regression Analysis of NSM FRP and FRP Wrapping Effects in Strengthened RC Columns

In this study, a simple regression-based approach is developed to accurately predict the axial load capacity of RC (Reinforced Concrete) columns strengthened using NSM (Near Surface Mounted) and hybrid FRP (Fiber Reinforced Polymer) methods. The model is trained and validated using a dataset of ۱۱۲ experimentally tested specimens collected from the literature, including ۲۲ data points from the authors' previous study and ۹۰ additional cases. Key input parameters include column dimensions (width and height), concrete compressive strength (mean ۳۵.۱ MPa), longitudinal steel ratio (mean ۱.۴۳%), transverse steel ratio, number and properties of NSM FRP and FRP jackets, and load eccentricity ratio (mean ۰.۱۸). The dataset was split into ۸۰% training and ۲۰% test sets. Three regression models were developed: (i) full linear regression, (ii) regression with statistically significant features, and (iii) stepwise regression. The stepwise regression model showed the best performance with a Mean Absolute Percentage Error (MAPE) of ۱۰.۸۵% and Root Mean Square Error (RMSE) of ۲۰۱.۹۱ kN on the test set. This approach eliminates complex trial-and-error procedures and offers a transparent, practical tool for engineers and researchers involved in strengthening RC columns.

کلیدواژه ها:

RC Column ، NSM FRP ، FRP jacket ، Axial Strengthening