Title:
Artificial Neural Networks for Prediction of Bond Strength between UHPC and FRP Reinforcing Bars
Author(s):
Ali Alatify and Yail J. Kim
Publication:
Symposium Paper
Volume:
363
Issue:
Appears on pages(s):
137-148
Keywords:
artificial neural network; bond; fiber reinforced polymer (FRP); machine learning; modeling; ultra-high performance concrete (UHPC)
DOI:
10.14359/51742111
Date:
7/1/2024
Abstract:
This paper presents the prediction of bond strength between ultra-high performance concrete (UHPC) and fiber reinforced polymer (FRP) reinforcing bars using an artificial neuronal network (ANN) approach. A large amount of datasets, consisting of 183 test specimens, are collected from literature and an ANN model is trained and validated. The ANN model includes six variable inputs (bar diameter, concrete cover, embedment length, fiber content, concrete strength, and rebar strength) and one output parameter (bond strength). The model performs better than other models excerpted from existing design guidelines and previously published papers. Follow-up studies are expected to examine the individual effects of the predefined input parameters on the bond strength of UHPC interfaced with FRP rebars.