Literature Synthesis -- Batch 02 Agent 5 (Files 361-400)¶
Generated: 2026-04-17 | 40 papers | Period: 2020-2021
Paper Extractions¶
| # | Author(s) | Year | Title | Core Finding | Method | Tags |
|---|---|---|---|---|---|---|
| 1 | Yang et al. | 2020 | Aero-hydro-servo-elastic coupling framework for FOWTs | F2A framework integrating FAST+AQWA shows excellent agreement with OpenFAST for FOWT dynamics | Coupled numerical simulation (F2A), verification against OpenFAST | FOWT, aero-hydro-servo-elastic, numerical-coupling |
| 2 | Zhang S. et al. | 2020 | Deep Learning for Bearing Fault Diagnostics -- Review | DL outperforms conventional ML in bearing fault feature extraction; bearing faults cause 30-40% of machine failures | Comprehensive survey; CWRU benchmark comparison of DL architectures | deep-learning, bearing-fault, SHM, condition-monitoring |
| 3 | Zhang P. et al. | 2020 | Random forest model for caisson foundation failure envelopes in sand | RF accurately learns failure mechanisms from FEM data and generalizes to new soils with minimal additional calibration | Random forest + coupled Lagrangian FEM-SPH (CLSPH-SIMSAND) | machine-learning, caisson-foundation, failure-envelope, sand |
| 4 | Zhang X. et al. | 2020 | Small-scale modelling of root-soil interaction under lateral loads | Shallow-wide root architecture gives higher anchorage; effective stress governs regardless of water condition | 1:20 scaled 3D-printed root models; centrifuge (20g) + 1g push-over tests | centrifuge, root-soil, scaling-laws, lateral-loading |
| 5 | Zhou et al. | 2020 | High-speed railway subgrade settlement in soft soil -- review | Post-construction settlement control in soft soil is critical; existing prediction methods have significant defects | Literature review of treatment methods and settlement prediction algorithms | soft-soil, settlement, railway, review |
| 6 | Zhu et al. | 2020 | Centrifuge modelling of tetrapod jacket foundation in soft soil | Jacket foundations accumulate deformation under cyclic lateral loading in soft soil; safe design must consider cumulative effects | Centrifuge modelling, lateral cyclic and monotonic loading | centrifuge, jacket-foundation, cyclic-loading, soft-soil |
| 7 | Zorzi et al. | 2020 | Reliability analysis of OWT foundations under lateral cyclic loading | Probabilistic framework quantifies SLS reliability for monopile rotation under extreme cyclic events using 3D FE | 3D FE with cyclic strain accumulation + reliability analysis | monopile, reliability, SLS, cyclic-loading, probabilistic |
| 8 | Ostlund et al. | 2020 | Model assumptions' influence on dynamic impedance of shallow foundations | Model assumptions (embedment, modulus variation, permanent load) can change static stiffness by >100% | FE parametric study with small-strain modulus reduction | dynamic-impedance, shallow-foundation, SSI, soil-dynamics |
| 9 | Kariyawasam | 2020 | Vibration-based bridge scour monitoring (PhD thesis, Cambridge) | Vibration-based monitoring can detect scour indirectly, avoiding fragile underwater equipment | Vibration-based indirect monitoring technique | SHM, scour, bridge, vibration-monitoring |
| 10 | Whyte | 2020 | Practical constitutive models in FEA of offshore foundations (DEng thesis, Oxford) | 3D FEA calibrates simplified 1D models for rapid foundation sizing; constitutive model choice is critical | 3D FEA with hypoplastic/critical-state models, macro-element calibration | FEA, constitutive-model, offshore-foundation, monopile |
| 11 | Samui et al. (eds.) | 2020 | Modeling in Geotechnical Engineering (book) | Comprehensive coverage of ML and numerical methods in geotechnical modelling | Edited volume: AI, FEM, and hybrid approaches | geotechnical-modelling, ML, book |
| 12 | Adedipe & Shafiee | 2021 | Economic assessment framework for OWF decommissioning | Turbine/foundation removal is the costliest stage; CBS approach identifies critical cost drivers | Cost breakdown structure (CBS), case study 500MW OWF | decommissioning, life-cycle-cost, offshore-wind |
| 13 | Ali et al. | 2021 | Offshore wind farm-grid integration review | PQ issues from PECs remain challenging; FACTS, DFIG control, ESS, and LVRT are key mitigation strategies | Comprehensive review of grid codes, PQ solutions | offshore-wind, grid-integration, power-quality, review |
| 14 | Almuqhim & Saeed | 2021 | ASD-SAENet: sparse autoencoder for autism detection via fMRI | Sparse autoencoder + DNN achieves improved ASD detection from fMRI data | Sparse autoencoder, deep neural network, fMRI | autoencoder, medical-imaging, deep-learning |
| 15 | Alves et al. | 2021 | PSD variability effect on SWCC predictions of coarse-grained materials | Cu influences residual suction prediction; D10 governs transition and residual stages of SWCC | Five SWCC prediction models; sieve, laser diffraction, image analysis | unsaturated-soil, SWCC, particle-size, variability |
| 16 | Anandan et al. | 2021 | CNN for soil texture property prediction from hyperspectral data | CNN with spatial interpolation of hyperspectral data predicts six soil properties (OC, CEC, N, pH, clay, sand) | 1D-CNN on hyperspectral arrays | CNN, soil-classification, hyperspectral, remote-sensing |
| 17 | Atar et al. | 2021 | First-order finite similitude in structural mechanics and earthquake engineering | Finite similitude theory provides countably infinite scaling alternatives beyond classical dimensional analysis | Zeroth- and first-order finite similitude; case studies on steel buildings with viscous dampers | scaling-laws, similitude, earthquake-engineering, physical-modelling |
| 18 | Basack et al. | 2021 | OWT power generation -- overview of research and developments | OWT technology still under development with significant unsolved problems in dynamics, foundations, and power efficiency | Literature review | offshore-wind, OWT, review |
| 19 | Bellan | 2021 | Statistical effects of frequency instability on power monitoring | Frequency instability introduces bias and standard deviation in harmonic power measurements via DFT | Analytical statistical characterization, DFT-based measurement model | power-measurement, frequency, signal-processing |
| 20 | Bendixen et al. | 2021 | Sand, gravel, and UN SDGs: conflicts, synergies, pathways | Sand/gravel extraction conflicts with multiple SDGs; governance frameworks needed for sustainable aggregate sourcing | Policy analysis mapping extraction impacts to 17 SDGs | sustainability, aggregate, SDG, environmental-policy |
| 21 | Bienen et al. | 2021 | Installation process effect on monopile lateral response | Impact-driven installation changes soil state, producing stiffer lateral response than wished-in-place piles | CEL large-deformation FEA with hypoplastic model + centrifuge validation | monopile, installation-effect, CEL, lateral-response, sand |
| 22 | Bortolotti et al. | 2021 | Land-based wind turbines with flexible rail-transportable blades | (File not found -- no extraction possible) | -- | -- |
| 23 | Bossuyt et al. | 2021 | Wake shape modulation for tilt and yaw misaligned wind turbines | Negative tilt (downward deflection) gives best wake recovery through increased curling and high-momentum downdraft | Stereo-PIV cross-plane measurements | wind-farm, wake-steering, tilt, yaw, PIV |
| 24 | Canet et al. | 2021 | On the scaling of wind turbine rotors | Scaling laws formulated for aeroservoelastic matching; even very large scaling factors can match key performance indicators | Constrained optimal aerodynamic/structural matching; 10MW reference scaled to 54m, 27m, 2.8m | scaling-laws, wind-turbine, aeroservoelastic, subscaling |
| 25 | Cerfontaine et al. | 2021 | Control of screw pile installation for offshore energy | Screw pile installation parameters (advance ratio) control performance; optimised installation improves capacity | Centrifuge testing, installation parameter control | screw-pile, installation, offshore, centrifuge |
| 26 | Chadebec et al. | 2021 | Data augmentation in HDLSS setting using geometry-based VAE | Riemannian-manifold VAE with normalizing flows significantly boosts classification in small-sample medical imaging | Geometry-based VAE, Riemannian metric learning, ADNI brain MRI | VAE, data-augmentation, small-sample, medical-imaging |
| 27 | Che et al. | 2021 | Transient wave-based anomaly detection in fluid pipes -- review | Five transient-wave methods reviewed (reflection, damping, FRF, time/frequency inversion); each has distinct applicability | Comprehensive review of wave-based pipe inspection methods | SHM, pipe-inspection, transient-wave, anomaly-detection |
| 28 | Chen C. et al. | 2021 | Efficient fatigue life prediction of OWTs using aerodynamic decoupling | Aerodynamic decoupling dramatically reduces computational cost for fatigue life prediction with acceptable accuracy | Decoupled aero-structural simulation | fatigue, OWT, aerodynamic-decoupling, computational-efficiency |
| 29 | Chen P. et al. | 2021 | Software-in-the-Loop + RL for FOWT dynamic response | Reinforcement learning combined with SIL simulation improves FOWT dynamic response prediction | Software-in-the-loop, reinforcement learning, OpenFAST | FOWT, reinforcement-learning, dynamic-response |
| 30 | Cooperman et al. | 2021 | Wind turbine blade material in the US: quantities, costs, EOL | 2.2M tons cumulative blade waste by 2050 (~1% landfill capacity by volume); current disposal costs are low | Spatially-resolved waste and cost estimation, 20-year lifetime assumption | blade-waste, circular-economy, end-of-life, composite |
| 31 | Cornel et al. | 2021 | Condition monitoring of roller bearings using acoustic emission | AE detects bearing damage ~4% earlier than vibration-based CMS; sub-surface damage signals detected ~50% before failure | AE sensors on downscaled bearing test rigs, comparison with vibration CMS | condition-monitoring, acoustic-emission, bearing, wind-turbine |
| 32 | Cross et al. | 2021 | Physics-informed ML for SHM | Grey-box models (physics + data) improve generalization beyond training regimes for SHM applications | Gaussian process regression with physics-informed kernels/mean functions | PIML, SHM, Gaussian-process, grey-box |
| 33 | Dhandapani & Varadarajan | 2021 | Multi-channel CNN for leaf disease and soil property prediction | MCNN with separate channels for leaf/soil images achieves 87.8% leaf disease and 90.4% soil property accuracy | Multi-channel CNN, Pearson correlation | CNN, soil-property, agriculture, multi-channel |
| 34 | Dirwai et al. | 2021 | Soil wetting geometry model under Moistube Irrigation | Buckingham pi-based empirical model estimates wetted dimensions with nRMSE 0.5-10% | Buckingham pi theorem, experimental calibration/validation | Buckingham-pi, irrigation, soil-wetting, empirical-model |
| 35 | Erinmwingbovo & La Mantia | 2021 | Instrument artefact correction in dynamic impedance spectra | Proposed method enables proper dynamic impedance above potentiostat bandwidth by correcting I/E converter artefacts | Transimpedance and stray capacitance estimation, correction algorithm | impedance-spectroscopy, artefact-correction, electrochemistry |
| 36 | Fan et al. | 2021 | Monopile installation effects on lateral response in sand (Part II) | Installation method (driven vs jacked vs wished-in-place) significantly affects subsequent lateral stiffness and capacity | CEL numerical analysis + centrifuge validation | monopile, installation, lateral-loading, sand, centrifuge |
| 37 | Galvin et al. | 2021 | Fast simulation of railway bridge dynamics with SSI | Sub-structuring with PML-based impedance functions + complex modal superposition matches FE-BE results at minimal cost | FE sub-structuring, PML, complex modal superposition | SSI, railway-bridge, dynamics, computational-efficiency |
| 38 | Ghimire et al. | 2021 | CNN-LSTM for streamflow prediction | CNN-LSTM outperforms standalone CNN, LSTM, DNN, and conventional AI; 84% predictions within 0.05 m3/s error | CNN feature extraction + LSTM sequential prediction | CNN-LSTM, hydrology, time-series, deep-learning |
| 39 | Gravett & Markou | 2021 | Wind turbine structures on soft clay: SFSI with battered RC piles | Optimum battered pile inclination is 10 degrees; failure localizes at tower base via local buckling | 3D hexahedral FE (superstructure + piles + soil), modal + pushover analysis | wind-turbine, SFSI, battered-piles, soft-clay, FEA |
| 40 | Grigoriu | 2021 | Linear Dynamical Systems (book, Cornell) | Textbook on SDOF/MDOF dynamics, frequency domain, numerical methods, shear beam models | Analytical + numerical methods for structural dynamics | structural-dynamics, textbook, shear-beam |
1. CONSENSUS¶
-
Installation effects matter for monopile design. Bienen (2021) and Fan (2021) independently demonstrate that impact-driven installation changes soil state and produces stiffer lateral response than idealized wished-in-place conditions. The field is converging on the view that ignoring installation effects leads to conservative or inaccurate foundation stiffness predictions.
-
Deep learning surpasses conventional ML for time-series and fault diagnostics. Zhang S. (2020) on bearing faults, Ghimire (2021) on streamflow, and Cross (2021) on SHM all confirm that DL architectures (CNN, LSTM, autoencoders) outperform traditional methods in feature extraction and prediction accuracy, particularly for nonlinear, non-stationary signals.
-
Soil-structure interaction is essential for dynamic structural assessment. Ostlund (2020), Galvin (2021), Gravett (2021), and Zorzi (2020) all show that SSI significantly alters dynamic impedance, modal properties, and structural response. Neglecting SSI can produce errors exceeding 100% in stiffness coefficients.
-
Centrifuge testing remains the gold standard for validating geotechnical numerical models. Zhang X. (2020), Zhu (2020), Bienen (2021), Fan (2021), and Cerfontaine (2021) all rely on centrifuge experiments to validate or calibrate numerical/analytical predictions of foundation behavior.
-
Coupled multi-physics simulation is necessary for OWT/FOWT design. Yang (2020), Chen C. (2021), and Chen P. (2021) demonstrate that aerodynamic, hydrodynamic, structural, and control coupling cannot be decoupled without careful justification, though decoupling can achieve computational savings when validated.
2. DEBATES¶
-
Wished-in-place vs. installation-modelled piles: The degree to which installation effects improve or degrade lateral capacity remains debated. Bienen (2021) shows stiffer response from driving, but pore-fluid effects during impact driving and vibro-driving remain unresolved.
-
Black-box vs. grey-box vs. white-box for SHM: Cross (2021) argues for physics-informed grey-box models that generalize beyond training data, while Zhang S. (2020) and Ghimire (2021) demonstrate strong purely data-driven results. The trade-off between interpretability/generalization and raw predictive power is unresolved.
-
Classical dimensional analysis vs. finite similitude: Atar (2021) challenges dimensional analysis as the sole scaling framework, proposing infinite alternative similitude orders. Whether this new theory is practical for routine scaled testing (versus classical Buckingham pi) remains under investigation (cf. Canet 2021, Zhang X. 2020).
-
Cyclic degradation models for SLS design: Zorzi (2020) highlights that soil response under cyclic loading involves nonlinearity, pore pressure, and stiffness changes, but the community lacks consensus on which constitutive framework best captures long-term cyclic accumulation for reliability-based design.
3. GAPS¶
-
Pore-fluid response during impact driving of monopiles is acknowledged as unresolved by both Bienen (2021) and Fan (2021). No validated coupled hydro-mechanical model of the driving process currently exists.
-
Long-term cyclic loading validation data for OWT foundations in soft clay are scarce. Gravett (2021) and Zhu (2020) rely on limited loading histories; multi-decadal field monitoring data do not exist.
-
Scaling laws for soil-foundation systems under dynamic/cyclic loading lack a unified framework. Atar (2021) offers theoretical advances but does not apply them to geotechnical problems. Canet (2021) addresses rotor scaling only; foundation scaling remains ad hoc.
-
Data augmentation for small geotechnical datasets. Chadebec (2021) demonstrates geometry-based VAE for medical imaging; equivalent techniques have not been applied to geotechnical data despite similar small-sample challenges (Zhang P. 2020 relies on numerical augmentation instead).
-
End-of-life strategies for composite blades beyond landfilling remain economically uncompetitive (Cooperman 2021). No cost-effective recycling technology has emerged at industrial scale.
-
Vibration-based scour monitoring (Kariyawasam 2020) needs field validation at scale and under varying flow conditions; lab-to-field transfer is incomplete.
4. METHODS¶
| Method Category | Papers | Notes |
|---|---|---|
| Centrifuge physical modelling | Zhang X., Zhu, Bienen, Fan, Cerfontaine | Standard for offshore/geotechnical validation; scales 20g-100g |
| Large-deformation FEA (CEL) | Bienen, Fan | Coupled Eulerian-Lagrangian for installation modelling |
| 3D FEA with constitutive models | Whyte, Gravett, Zorzi, Ostlund | Hypoplastic, critical-state, small-strain models |
| Machine learning (RF, CNN, LSTM) | Zhang P., Zhang S., Anandan, Ghimire, Dhandapani | Supervised learning on numerical/experimental databases |
| Physics-informed ML | Cross | Gaussian process regression with physics kernels |
| Autoencoders/VAE | Almuqhim, Chadebec | Sparse AE for classification; geometry-based VAE for augmentation |
| Reliability/probabilistic | Zorzi | Monte Carlo + FE for SLS reliability index |
| Sub-structuring / impedance functions | Ostlund, Galvin | Frequency-dependent stiffness/damping for SSI |
| Buckingham pi / dimensional analysis | Dirwai, Atar, Canet | Classical and extended scaling frameworks |
| Coupled aero-hydro-servo-elastic | Yang, Chen C., Chen P. | FAST+AQWA, decoupled approaches, RL-augmented |
| Acoustic emission monitoring | Cornel | AE for pre-damage detection in bearings |
| Stereo-PIV | Bossuyt | Wake flow measurement for tilt/yaw studies |
5. BENCHMARKS¶
| Benchmark | Used By | Domain |
|---|---|---|
| CWRU bearing dataset | Zhang S. (2020) | Bearing fault diagnostics DL comparison |
| NREL 5MW / OC3-Hywind spar | Yang (2020) | FOWT aero-hydro-servo-elastic validation |
| NREL 10MW reference turbine | Zorzi (2020), Canet (2021) | Monopile reliability; rotor scaling |
| Brisbane River / Teewah Creek flow data | Ghimire (2021) | Streamflow CNN-LSTM prediction |
| ADNI brain MRI database | Chadebec (2021) | VAE data augmentation for medical imaging |
| Karlsruhe/UWA sand databases | Bienen (2021), Fan (2021) | CEL installation modelling validation |
| WindAfrica geotechnical site data | Gravett (2021) | Soft clay SFSI pile optimization |
| 500MW baseline OWF | Adedipe (2021) | Decommissioning cost estimation |