Abstract
This paper proposes a fault detection method for blade pitch systems of floating wind turbines using transformerbased deep-learning models. Transformers leverage self-attention mechanisms, efficiently process time-series data, and capture long-term dependencies more effectively than traditional recurrent neural networks (RNNs). The model was trained using normal operational data to detect anomalies through high reconstruction losses when encountering abnormal data. In this study, various fault conditions in a blade pitch system, including environmental load cases, were simulated using a detailed model of a spar-type floating wind turbine, the data collected from these simulations were used to train and test the transformer models. The model demonstrated superior fault-detection capabilities with high accuracy, precision, recall, and F1 scores. The results show that the proposed method successfully identifies faults and achieves high-performance metrics, outperforming existing traditional multi-layer perceptron (MLP) models and long short-term memory-autoencoder (LSTM-AE) models. This study
highlights the potential of transformer models for real-time fault detection in wind turbines, contributing to more advanced condition-monitoring systems with minimal human intervention.
Key Words
blade pitch system; fault detection; floating wind turbine; prognostics and health management; sequential data; transformer
Address
Seongpil Cho, Sang-Woo Kim: Department of Aeronautical and Astronautical Engineering, Korea Aerospace University, 76 Gonghangdaehak-ro, Deokyang-gu, Goyang, Gyeonggi 10540, Republic of Korea
Hyo-Jin Kim: Department of Korean Medical Science, Kyung Hee University, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea
Abstract
Composite beams of steel and concrete strengthened with fiber-reinforced polymers (FRP) may exhibit considerably
enhanced flexural behaviour, but the combination of three materials with different characteristics and the various possible failure mechanisms that may govern performance make their analysis quite demanding. Previous studies provided significant insights into this problem and several methods were proposed for calculating flexural stiffness and strength, but these studies are restricted to the single member level of a simply supported composite beam section. However, the problem considerably changes when the beam is part of a frame system due to the degree of continuity provided by the surrounding structure, which represents the most common situation in practice. This paper explores the behaviour of semi-continuous FRP-strengthened composite beams, by considering the response characteristics of their end connections and their effects on overall performance. A novel analytical model is derived, which enables a step-by-step representation of the nonlinear relationship between an
incremental mid-span design bending moment and corresponding connection rotations. After verification against finite element analyses, a parametric study is conducted which shows that the substantially increased bending moment resistance of FRPstrengthened composite beams can hardly be fully utilized due to a deficiency of corresponding large deformation capacity available in the connections. The extent to which the presence FRP strengthening can be exploited to enhance the beam flexural response depends on the interplay between various structural parameters, including the connection rotation capacity, the beam
span, and the FRP modulus of elasticity and ultimate strength.
Address
Panagiotis M. Stylianidis: Department of Civil Engineering, Neapolis University Pafos, Paphos 8042, Cyprus
Michael F. Petrou: Department of Civil and Environmental Engineering, University of Cyprus, Nicosia 2109, Cyprus
Abstract
Previous research has identified inadequate flexibility in concrete pavements due to the use of high-strength concrete
mixtures. This research investigates whether this problem can be addressed by partially replacing some fine and coarse
aggregate components with waste rubber from shredded tires, the safe disposal of which otherwise is a major environmental concern. Using finite element software ABAQUS, this study analyses 3D pavement model behavior in terms of internal stress development and deflection at critical load points. This analysis is carried out for concrete slabs of differing waste rubber proportions and varying thicknesses. Results show that the maximum tensile stress is reduced, and maximum deflection is increased as the rubber content in pavement concrete slab is increased. The stresses and deflection of concrete pavement slab are reduced as the thickness of the slab is increased. The influence of increasing the base coarse modulus is significant in terms of reduction in tensile stress development. However, the reduction in deflection is found to be relatively marginal, especially in low-percentage rubberized pavement concrete slabs.
Key Words
3D pavement model; chip rubber; concrete pavements; crumb rubber; finite element analysis; waste shredded tires
Address
Amin Hamdi, Khatib Zada Farhan and Sohaib Gutub: Civil and Environmental Engineering Department, King Abdulaziz University, Jeddah, Saudi Arabia
Abstract
This paper intends to analyze the nonlinear forced vibrations of functionally graded material (FGM) beams with
initial geometrical defects in hygro-thermal ambiences. For this purpose, we assume that the correlation properties of the material alter along the thickness direction in succession and the surface of the beam is subjected to humid and thermal loads. Based on the Euler Bernoulli beam theory and geometrical non-linearity, we use the Hamiltonian principle to formulate a theoretical model with consideration of the hygrothermal effects. Galerkin's technique has been proposed for the control equations of discrete systems. The non-linear primary resonances are acquired by applying the modified Lindstedt-Poincare
method (MLP). Verify the reliability of the data obtained through comparison with literature. The non-linear resonance response is reflected by amplitude-frequency response curves. The numerical results indicate that the resonances of FGM beams include three non-linear characteristics, namely hard springs, soft springs and soft-hard spring types. The response modalities of the structure may transform between those non-linear characteristics when material properties, spring coefficients, geometric defect
values, temperature-humidity loads and even the external stimulus generate variations.
Abstract
In this study, the cyclic behavior of Bar Damper (BD) and its effect on the seismic performance of the steel frame was investigated using numerical and analytical methods. Initially, the calibrated model was used to conduct parametric studies on the cyclic behavior of the damper. The purpose of parametric studies was to provide equations for calculating effective and elastic stiffness, ultimate strength, and energy dissipation using its diameter and height. The impact of BD on the steel frame was examined in the second section of the research. In this section, studies were conducted using pushover analysis to investigate the impact of BD on the elastic stiffness, energy absorption, ductility, and strength of the frame. The results demonstrated that increasing the height of the BDs resulted in higher energy dissipation. However, reducing the height and increasing the diameter increased effective stiffness, yield strength, and elastic stiffness. The EVDR results showed that the diameter of the damper has a negligible effect on it, and its value increases with the decrease in height. In the best case, the addition of BD causes a 23% increase in energy dissipation and a 60% increase in frame ductility.
Key Words
bar damper; cyclic behavior; numerical method; steel frame; yielding damper
Address
Kambiz Cheraghi: Department of Civil Engineering, Faculty of Engineering, Razi University, Kermanshah, Iran
Mehrzad TahamouliRoudsari: Department of Civil Engineering, Kermanshah Branch, Islamic Azad University, Kermanshah, Iran
Reza Aghayari: Department of Civil Engineering, Faculty of Engineering, Razi University, Kermanshah, Iran
Kaveh Cheraghi: Department of Mechanical Engineering, Engineering Faculty, Razi University, Kermanshah, Iran
Abstract
Metal-based materials used in ships are built by welding plates and profiles of various sizes and shapes together. Although various methods are currently used during the production of ships, studies are ongoing on alternative welding methods. When alternative methods are examined, it is seen that friction stir welding (FSW) is advantageous in applying plate-type materials and obtaining high mechanical properties after application. In this study, FSW was applied to the steel used in ships, and after the application, hardness, tensile, and bending tests were performed, and mechanical properties were determined. Afterward, the bending test results, which are of great importance for the formability of welded structures, were transferred to finite element analysis (FEA) and multilayer perceptron (MLP) models, and the data obtained in these models were mutually analyzed with the mechanical test data. As a result of the analyses, it was determined that models with appropriate results obtained with experimental data could be created after both FEA and MLP, and thus the bending behavior of welded structures could be determined without the need for experimental data.
Key Words
finite element analysis approach; friction stir welding; high-strength shipbuilding steel; multilayer perceptron (MLP)
Address
Dursun Murat Sekban: Department of Marine Engineering Operations, Karadeniz Technical University, 61530, Trabzon, Turkey; WMS Engineering Services Industry Trade Limited Company, 61080, Trabzon, Turkey
Ecren Uzun Yaylaci: Faculty of Fisheries, Recep Tayyip Erdogan University, 53100, Rize, Turkey; Faculty of Engineering and Architecture, Recep Tayyip Erdogan University, 53100, Rize, Turkey
Mehmet Emin Özdemir: Department of Civil Engineering, Cankiri Karatekin University, 18100, Çankiri, Turkey
Murat Yaylaci: Department of Civil Engineering, Recep Tayyip Erdogan University, 53100, Rize, Turkey; Faculty of Turgut K
Abstract
In this study, a unified framework that integrates pre- and post-earthquake assessments of buildings was proposed to
enhance urban disaster preparedness through the coordination of pre- and post- earthquake efforts. Within this framework, a case study based on the 2023 Kahramanmaraş Earthquake was performed comparing the distribution of seismic risk prioritization for 117 reinforced concrete buildings with their actual damage states observed during post-earthquake field inspections. In order to conduct pre-earthquake evaluation process, street-level images were employed using two different rapid visual screening methods. With the use of generated geospatial database enabling the efficient and reliable transmission of the data between both stages of the assessment procedures, the alignment and validation of pre- and post-earthquake evaluations of the buildings were
achieved enhancing the coordination of seismic risk management strategies. By implementing the proposed joint framework in this study, an extensive seismic vulnerability evaluation on an urban scale could be achieved by optimizing the computational demands, cost and time required for the strategic planning activities.
Address
Ayşe E. Özsoy Özbay, Işil Sanri Karapinar: Department of Civil Engineering, Maltepe University, Marmara Eğitim Köyü, 34857, İstanbul, Turkey
Hüseyin C. Ünen: Yer Çizenler Mapping for Everyone Association, 34722, İstanbul, Turkey
Abstract
Recently, the phenomenon of disproportionate structural failure caused by blast load has grown more common in the field of engineering design. Blast-resistant analyses and designs have been developed by many structural techniques and methodologies to forecast the loads produced by a high explosive charge on structures with complicated geometry. These techniques are based on a good understanding of blast phenomena to analyze structures exposed to blast load. This paper provides a current state-of-the-art review of blast prediction and simulation methods to predict the design blast loads that are used to assess the structural response and damage level to an existing or new building. The damage criteria from the general design approach relevant to civil design applications in forecasting blast loads as well as structural system responses will be provided. Identifying the structures' expected damage class would aid in providing extra reinforcing or strengthening for damaged elements to meet the acceptance criteria or minimize damage by a suitable blast mitigation strategy. Based on identifying the damage class expected of a structure subjected to an explosion, blast mitigation strategies could be used to minimize damage and maximize the ability of the structure to function even after the explosion.
Address
Tarek Sharaf, Sara Ismail, Mohamed Elghandour and Ahmed Turk: Department of Civil Engineering, Faculty of Engineering, Port Said University, Galaa Street, City of Port-Fouad, Port Said, PO 4252, Egypt
open access
