Skip to main navigation Skip to search Skip to main content

Data-driven analysis of frictional behavior between CFRP lamella and wave-shaped steel clamp plates under normal compression

  • Jiazhan Xie
  • , Yue Liu*
  • , Chong Wang
  • , Mudassir Iqbal
  • , Xin Qiang
  • , Tao Wu
  • , Lik ho Tam
  • *Corresponding author for this work
  • Southeast University, Nanjing
  • University of Science and Technology Beijing
  • University of Engineering and Technology, Peshawar
  • Beihang University
  • Fraunhofer Institute for Ceramic Technologies and Systems

Research output: Contribution to journal › Article › peer-review

Abstract

Due to supreme mechanical properties and corrosion resistance, carbon fiber reinforced polymer (CFRP) has been utilized in combination with metal to fabricate anchor systems. In these anchor systems, the coefficient of friction of the interface is a crucial design criterion as it determines the mechanical performance of the CFRP-steel element. In this work, a novel anchor system was designed where carbon fiber composites lamella is clamped by a pair of wave-shaped steel plates, with a focus on the frictional behavior between the lamella and wave-shaped steel clamp plate. Specifically, the coefficient of friction of the interface was experimentally measured as a function of the variable thickness of CFRP lamella, the wave peak (height), and the compression force of the steel clamp. It was found that for the wave peak, a height of 2 mm was the optimum value for CFRP with a thickness of 1.2, 2, and 3 mm. Meanwhile, high compression of the clamp yields an early failure of the CFRP at that region and a compression force of 105 kN is measured as the threshold. The failure occurred at lower breaking force because the CFRP lamella was gradually scraped off and its transverse section (thickness) was dropped to a smaller cross-section during friction testing at higher contact forces (70 kN and 105 kN). The developed nonlinear ANN and XGBoost models performed robustly for the training (RA = 0.980, RX = 0.966), validation (RA = 0.963, RX = 0.912), and test data (RA = 0.911, RX = 0.894). Based on the results, it is demonstrated that thickness and wave peak manifested an inverse relationship, and the designed parameters of the anchor system can be optimized for mechanical performance according to practical needs.

Original languageEnglish
Article number112763
JournalJournal of Building Engineering
Volume107
DOIs
StatePublished - 1 Aug 2025

Keywords

  • Artificial neural network
  • CFRP lamella
  • Coefficient of friction
  • Data-driven estimation
  • Wave-shaped steel clamp plates

Fingerprint

Dive into the research topics of 'Data-driven analysis of frictional behavior between CFRP lamella and wave-shaped steel clamp plates under normal compression'. Together they form a unique fingerprint.

Cite this