Existing Operator Inference models accurately predict ship motions within sampled sea states but their accuracy may degrade under unseen wave conditions because a single fixed operator cannot fully represent the variation of system dynamics across different sea states. Building on a previously developed dual-memory non-Markovian Operator Inference framework, this study proposes a sea-state-conditioned local expert Operator Inference method for cross-period prediction of ship motions in irregular waves. The inferred evolution operator is parameterized by peak frequency and significant wave height, enabling the operator to adapt to different sea states while retaining the interpretable dual-memory representation of wave-input history and latent response memory. To reduce the mismatch between training and target conditions, a local expert strategy is introduced by selecting neighboring peak-period groups for operator identification. The framework is evaluated using CFD-generated motion data of a KCS container ship under 24 irregular-wave conditions. A leave-one-peak-period-out strategy is adopted to assess generalization to unseen wave periods. and four model variants are compared. Results show that the proposed local dynamic model achieves the best prediction accuracy, especially in short-period extrapolation cases where fixed-operator models show larger errors.

Digital Twin (DT) modelling is a crucial asset to provide structural health monitoring (SHM) for Floating Offshore Wind Turbine (FOWT), which aims to enhance the structural integrity and minimize the operation and maintenance expenditure (OPEX) under severe ocean environment. DT models are required to capture the important dynamics and run in real-time. Data-driven DT models are beneficial to produce highly nonlinear dynamic models in real-time and the existing models are mostly based on traditional machine learning techniques, which generate the surrogate models in a "black box" manner, the input and output data are blindly fed. This hinders the generalizability and adaptability to unknown scenarios. To address the SHM practical issues and overcome the limitation of existing data-driven DT models, we present the first Explainable Artificial Intelligence (XAI) framework for structural dynamics of FOWT with the expressive power of the novel Heterogeneous Spatial Temporal Graph Neural Network. The aero-hydro-servo coupled simulation results are generated from the software QBlade. The semi-submersible OC4 DeepCwind FOWT is analysed with the circumstances of above rated wind speed with different wind directions, rough sea state and sea current condition. This novel XAI framework can interpret the complicated aero-hydrodynamic coupling of FOWT, the imbalance loads created from the misalignment of wind inflow directions against the wave and current direction are investigated. Remarkably this is the first XAI illustrating the wind-wave load directionality with latent chaotic structural dynamics interaction in time domain. Time domain analysis is more prominent to dynamic internal stress analysis and nonlinear effect for fatigue analysis than frequency domain analysis, especially this new XAI model can include the secondorder hydrodynamics in real-time and provide more precise remaining useful life predictions for high-cycle fatigue in the subsequent stages.

Floating Offshore Wind Turbines (FOWT) provided new potential in harvesting wind energy in far offshore deep-sea regions and contributed to the world decarbonization Net-Zero target. Providing structural health monitoring (SHM) is crucial for ensuring the structural integrity of FOWT in lifecycle. However, the SHM is technically challenging with high Operational and Maintenance Expenditure (OPEX). Recently, Digital Twin (DT) and advanced sensor technologies offer alternative solutions to provide effective strategy in SHM remotely. Data-driven DT with deep learning models can formulate highly nonlinear dynamics systems. Yet, these existing models only perform the “black box” prediction without explicitly modeling the spatial-temporal relationship and consider only homogenous loading exerted in contrast to the complicated loading combination of FOWT with wind, wave and sea current. To address the existing modelling limitations, a new Graph Neural Network (GNN)-Encoder-Decoder-Long Short-Term Memory (LSTM) surrogate model of FOWT is presented in this work, which can perform 50 times faster than the real-time of simulation data set with accurate prediction of wind turbine tower bottom forces in the dominant dynamic modes force-aft and side-side directions. The training data is based on the software QBlade simulation and focuses on the OC4 5MW DeepCwind FOWT structure. A holistic quantitative analysis is carried out to validate the tractable latent space vectors for this complex FOWT system.

The predictive modeling of dynamic wake flow for floating offshore wind turbines presents significant challenges. In this paper, we propose a predictive data-driven model of wake flow for floating offshore wind turbines. The data-driven model is based on a high-order variant of the dynamic mode decomposition approach, incorporating a network of optimized data probe points and a discrete empirical interpolation method in preprocessing. A high-fidelity computational fluid dynamics model is utilized to study the wake flow pattern and generate data for our data-driven model. Our data-driven model can quickly forecast wake dynamics based on a limited number of physical state inputs, which are measured at a few probe points behind the wind turbine. The results show that our model can give a highly accurate future forecast of wake dynamics behind the wind turbine for up to 30 s and a moderately accurate forecast for up to 100 s and beyond. The frequency spectrum of our predictions also agrees well with the benchmark solutions.

Propeller boss cap fin (PBCF) is an energy- saving device that enhances the efficiency of marine vessels by mitigating propeller hub vortex and improving propulsion performance. In this work, a design procedure of the PBCF is introduced. It involves airfoil and fin shape optimization as well as PBCF performance evaluation. The airfoil optimization is performed under given ship advance ratio and propeller rotating speed, using an engineering tool and surrogate-based optimization to achieve maximum net thrust. The optimized airfoil demonstrates improved performance compared to the baseline design. The pitch angle describing the orientation of the PBCF is then optimized through iterative adjustments. The adjustment is guided by performance data obtained from computational fluid dynamics (CFD) simulations. The thrust and torque on the fin, cap and propeller are analyzed for better understanding of the force contribution from each component. Both straight fin for constant pitch angle and twisted fin with linearly decreased pitch angle are investigated. Moreover, simulations are conducted to evaluate the performance with a convergent cap as a replacement of the straight cap. CFD simulations show that the efficiency of the propeller equipped with the optimized PBCF and a convergent cap can increase by nearly 1% compared to the propeller without PBCF. The increase in propeller efficiency with optimized straight fins is higher than that achieved with twisted fins. With this design optimization framework, propeller performance can be improved, resulting in better fuel efficiency and cost savings

Structural integrity of Floating Offshore Wind Turbine (FOWT) is the prominent factor that can affect the whole lifecycle of offshore wind project. In practice, structural health monitoring of FOWT is technically challenging under the harsh sea conditions, especially the limitation of sensor placement for internal force measurement. Therefore, Digital Twin (DT) is an important asset that allows real-time remote monitoring and reflects the real condition on site. Recent DT development based on data-driven approach has been explored in offshore structures force monitoring and forecasting. The existing deep learning-based models lack the expressiveness of the geometrical, structural and material properties. To address the problems in the current industry and the limitation of existing simulation approach, a novel finite element surrogate model of FOWT based on Physics-Guided Graph Neural Network (PhyGGNN) is presented. The aero-hydro-servo coupled simulation results are generated from the software QBlade. The semi-submersible OC4 DeepCwind FOWT is considered under the above rated wind speed, rough sea state and water current condition according to metocean of West of Barra, Scotland. In addition, we present the first application of spatial-temporal GNN approach for solving offshore structural dynamics and accurate real-time prediction of wind turbine tower forces under complex wind, wave and current condition for FOWT. Internal forces prediction can allow the remaining useful life calculation in the next stage of fatigue analysis.

Abstract Offshore wind energy plays a key role in the transition to renewable energy sources. With the growing interest and installations of offshore wind farms, a deeper understanding of how extreme wind conditions in the coastal area impact power production becomes essential. According to the working regimes of a wind turbine, wind energy production can be affected or disrupted by extreme atmospheric events particularly when facing specific thresholds such as below the cut-in wind speed, reaching the rated output speed and exceeding the cut-off wind speed. Given that these wind thresholds might all occur during extreme wind events, accurately calculating power production capacity during a coastal extreme atmospheric event becomes critical to assess wind farm availability. In order to evaluate the power production at a proposed offshore wind farm site during an extreme wind event, this study utilized a coupled atmosphere-wave-ocean modelling system (COAWST) comprising fully coupled configuration (ROMS-SWAN-WRF) to represent the complex meteo-oceanographic conditions along US east coast area: Gulf of Maine. The selection of the fully coupled configuration was based on its superior performance, validated through comparison with various buoy observations. Subsequently, the study compared the gridded wind power production from different wind turbine configurations during an extreme winter storm event. Power production was computed separately based on each turbine model’s power curve and thrust curve. Assessing the spatial and temporal variations in power production capacity among different turbine model configurations during this extreme winter event in this study not only advances the understanding of wind assessment over the Gulf of Maine but also provides invaluable insights for researchers into advantages of employing more advanced meteo-oceanographic modelling system, such as COAWST, in offshore wind farm industry.

In the field of fluid mechanics, it is a potential consensus that nonlinear dimensionality reduction (DR) techniques outperform linear methods. However, this conclusion has been obtained based on simple fluid phenomena and an incomplete evaluation system of dimensionality reduction algorithms. In this study, we use an improved evaluation system of DR methods to compare and evaluate the performance of four DR methods, including two linear techniques: Principal Component Analysis (PCA) and Independent Component Analysis (ICA), and two non-linear techniques: Isometric Mapping (ISOMAP) and Locally Linear Embedding (LLE). The four methods are applied to analyze a complex hydrodynamic flow field with cavitation by considering the joint features of multiple variables. Results show that LLE can capture redundant features that do not contribute to understanding the characteristics of flow fields, while ISOMAP is more suitable for handling datasets with multiple scales than LLE. PCA and ISOMAP can successfully capture the characteristics and evolution of flow fields. In addition, 3D supplementary information can assist ICA in improving the problem of identifying unstable flow field states.

Respiratory papilloma is a relatively common benign tumor of the respiratory tract, and a few patients may develop malignant changes. The disease has an insidious onset and lacks specific clinical manifestations, and its manifestations are closely related to the growth mode, location and size of the tumor. It can involve multiple parts, such as the larynx, trachea, bronchus, and lung parenchyma, which cause coughing, hoarseness, dysphonia, and, in severe cases, may lead to obstruction of the respiratory tract. At present, the treatment of respiratory papilloma lacks standardization, and there is no effective method to cure the disease. Surgery remains the main treatment for alleviating patients' symptoms and preventing airway obstruction. However, due to the high recurrence rate of respiratory papilloma, multiple surgeries are often needed, which reduces the quality of life of patients and increases their disease burden and economic burden. Bevacizumab, a vascular endothelial growth factor-binding antibody inhibitor, is a promising adjuvant treatment modality that shows good potential for reducing symptoms and the frequency of surgery. This article aimed to review the efficacy and safety of bevacizumab for the treatment of respiratory papilloma and discuss the differences and efficacy of the systemic application and intralesional injection of bevacizumab for the treatment of respiratory papilloma.

Objective: To construct and characterize conditional Src homology region 2 protein tyrosine phosphatase 1 (SHP-1) knockout mice in airway epithelial cells and to observe the effect of defective SHP-1 expression in airway epithelial cells on the emphysema phenotype in chronic obstructive pulmonary disease (COPD). Methods: To detect the expression of SHP-1 in the airway epithelium of COPD patients. CRISPR/Cas9 technology was used to construct SHP-1flox/flox transgenic mice, which were mated with airway epithelial Clara protein 10-cyclase recombinase and estrogen receptor fusion transgenic mice (CC10-CreER+/+), and after intraperitoneal injection of tamoxifen, airway epithelial SHP-1 knockout mice were obtained (SHP-1flox/floxCC10-CreER+/-, SHP-1Δ/Δ). Mouse tail and lung tissue DNA was extracted and PCR amplified to discriminate the genotype of the mice; the knockout effect of SHP-1 gene in airway epithelial cells was verified by qRT-PCR, Western blotting, and immunofluorescence. In addition, an emphysema mouse model was constructed using elastase to assess the severity of emphysema in each group of mice. Results: Airway epithelial SHP-1 was significantly downregulated in COPD patients. Genotyping confirmed that SHP-1Δ/Δ mice expressed CC10-CreER and SHP-1-flox. After tamoxifen induction, we demonstrated the absence of SHP-1 protein expression in airway epithelial cells of SHP-1Δ/Δ mice at the DNA, RNA, and protein levels, indicating that airway epithelial cell-specific SHP-1 knockout mice had been successfully constructed. In the emphysema animal model, SHP-1Δ/Δ mice had a more severe emphysema phenotype compared with the control group, which was manifested by disorganization of alveolar structure in lung tissue and rupture and fusion of alveolar walls to form pulmonary alveoli. Conclusions: The present study successfully established and characterized the SHP-1 knockout mouse model of airway epithelial cells, which provides a new experimental tool for the in-depth elucidation of the role of SHP-1 in the emphysema process of COPD and its mechanism.

FPSO is one of the largest vessels in offshore industry, the complexities of its topside structures pose a big challenge in reliable prediction of wind loads both with CFD modeling and in wind tunnel test. In this study, we have conducted detailed CFD studies of wind flow over a generic FPSO model at both full scale and model scale of 1:400, including the impact of different wind velocities and wind profiles. In the model scale, the simulations are performed at the constant wind velocity of 10m/s and 20m/s, for all wind headings between 0 to 360 degrees with a step of 10 degrees. The calculated wind loads in the form of coefficients of three force components and three moment components are validated against experimental measurements. In addition to the overall wind loads on the FPSO, the static surface pressures at 28 sensor locations were also measured during the wind tunnel test. In general, they are compared well with CFD simulation results. For the full-scale model, the CFD simulations are performed at different design wind speeds, the results confirmed the scalability of wind loads at the model scale. The impact of wind profile shape is addressed in this paper as well.

Nowadays, most prognostic models heavily rely on the complete and intact historical degradation signals to identify underlying deteriorated trends for predicting the lifetime of the engineering system. However, in real-world scenarios, these degradation signals are always collected with inconsistent distributions and incomplete observations, which compromises the ability to establish precise degradation models. Therefore, we have formulated an adjustable functional lifetime regression model with the real-time prognostic capability to tackle distribution shifts and incomplete data. Firstly, feasible degradation curves and informative features are identified through a functional perspective. Consequently, the relationships between the represented features and the lifetime labels are formulated by the innovative regression model with an adjustable functional basis. Finally, by leveraging real-time signals, our method can refine and update the time-to-failure (TTF) results. The experimental results significantly demonstrate the prognostic robustness, evaluation precision, and application prospects of the proposed approach.

As the Proper Orthogonal Decomposition (POD) based Discrete Empirical Interpolation Method (DEIM) constructs specially selected interpolation indices that define an interpolation-based projection without carrying out the orthogonal projection as the POD-based Reduced Order Method (ROM) does, we apply the DEIM method to reconstruct and predict the wind load on an FPSO in this study. High-fidelity CFD simulation is adopted to collect snapshots for the reduced-order model. Effects of the number of DEIM selected interpolation points are tested. Testing results indicate that DEIM can predict the wind load on the FPSO with high accuracy even with only 5 interpolation points. Considering the limited installation locations over the scaled models in wind tunnels, DEIM’s feature of interpolation indices selection for an optimal subspace approximation could be helpful to determine the optimal sensor positions, and consequently to offer guidance for sensor positioning for wind tunnel tests and improve the prediction accuracy of the ROM-based data assimilation models.

This paper presents a comprehensive study of wind loads on various offshore platforms using both wind tunnel experiments and CFD at model scales. The platforms studied include an LNGC with simple topside geometry, a FPSO with complex porous blocks, and a JUP with mixed simple blocks and tall truss structures. The 3D-printing is used to fabricate experimental models to ensure consistency with CFD models. Sensitivity studies of mesh types, mesh independence, turbulence models, wind profiles and wind speed independence are conducted to achieve highly accurate and reliable simulation results. Comparisons between the CFD and experimental results of six loading components and surface pressure measurements at over 20 sensor locations for each platform show good agreement across various wind directions and geometries. The robustness of the experiments and CFD models can serve as a reference for future offshore wind loading studies. Detailed examinations of the wind fields surrounding the platforms and the effects of porosity levels on wind loading reveal unique internal flow structures that differentiate among the structures and provide explanations for variations in loading and discrepancies. These findings contribute to a better understanding of wind fields and loading on the superstructures of offshore platforms and can inform the design and operation of similar structures in the future.

Clustering applied to unsteady flow fields can simplify flow field data and partition the flow field into regions of interest. Unfortunately, these areas are often unexplored when applied to complex fluid mechanics problems because multivariate data are difficult to express, and the relationships between flow field snapshots in a time series are difficult to preserve. In this paper, we use joint principal component analysis (JPCA) and fusion principal component analysis (FPCA) to process multivariate data to obtain the static and dynamic characteristics of the cavitation flow field. Based on the static characteristics of the flow field, we use the K-means algorithm and cohesive hierarchical clustering to obtain static flow field segmentation at different levels. Based on the dynamic characteristics of the flow field, we use the proposed time series K-means (TK-means) algorithm and cohesive hierarchical clustering to obtain dynamic flow field segmentation at different levels. The results show that JPCA or FPCA is effective in expressing multivariate features. Static flow field segmentation can obtain time-invariant, physically related structures of unsteady flow. Dynamic flow field segmentation can obtain time-varying, physically related structures of unsteady flow.

Performing numerical simulations of a scaled-down experimental setup of a floating offshore wind turbine allows for a two-way validation between the two approaches. Furthermore, additional insight in the physical behavior of the system can be obtained by tuning numerical results on experimental measurements, and exploring the results. Numerical simulations of a 10MW floating offshore wind turbine model in a wave basin are performed at model scale using computational fluid dynamics. The performance-scaled turbine is designed to match the Froude-scaled thrust force in a Froude-scaled wind field. Initial results show a discrepancy in calculated turbine thrust and torque results compared with experimental results. Two main challenges are identified: (1) selecting a numerical approach appropriate for the low-Reynolds number flow, and (2) accurately modelling the environmental conditions in the basin. In this paper, a numerical sensitivity study is carried out by varying systematically the turbulence models and inflow conditions. A standard k − ε turbulence model is used, as well as a more extensive γ – Reθ turbulence transition model. Different inflow conditions are set up to model the turbulent jet wind field in the wave basin. It is found that the k-ε turbulence model is unsuitable to match the model test results, while satisfying results are obtained using the γ – Reθ model. Furthermore, it is seen that the turbulent jet inflow is represented well by both a vertical power law profile and a radial profile fitted to wind field measurements, while uniform and tabulated inflow conditions are a poor representation of the experimental conditions. It is concluded that effects from surface roughness, the Reynolds number, the inflow velocity and turbulence distribution must be included in the numerical evaluation.

The Late Permian coal-accumulating area in eastern Yunnan Province and western Guizhou Province, especially Xuanwei Area, has extremely high incidence rate and mortality of lung cancer in the world. Genomic research on the lung cancer patients found a unique incidence mode specified to the environment. Carcinogens in C1 coal burning may be related to the high incidence of uncommon mutations G719X, S768I plus T790M, and G719X plus S768I, which were significantly higher than those in other regions.

In this study, the gappy Proper Orthogonal Decomposition (POD) method is adopted to fuse wind-tunnel measured pressure and computational fluid dynamics (CFD) simulation results to reconstruct the pressure field and calculate the force coefficients on a marine vessel. The technique is demonstrated for wind load evaluations on the LNG carrier GALEA. With 24 pressure sensor data from wind tunnel tests, the pressure distributions on the whole vessel surface are reconstructed successfully, and the force coefficients obtained from the gappy POD show a reasonable agreement with the wind-tunnel measured results and those obtained from CFD simulations. In addition, sensitivity studies have been carried out to determine the minimum sensor number requirement and sensor deployment strategies for gappy POD to achieve high accurate force coefficient evaluations.

In this work, the mechanisms responsible for the premature fatigue fracture of four suspension springs during road test were investigated. Failure analysis combined with finite element method (FEM) and stress corrosion cracking (SCC) experiments were conducted. The results indicate that the corrosion and wear damage occur on the failed spring surface caused by the stone chipping and acid rain. The unexpected failure of suspension springs was caused by the presence of surface cracks formed by the corrosion fatigue cracking (CFC) mechanism. The fatigue strength of spring reduces to as low as 420 MPa due to the presence of these surface cracks, which is much lower than the service stress. As a result, the fatigue crack propagates quickly without the help of corrosion and leading to the final rupture.