DOMAIN MAP 1: Geotechnical Foundations for Offshore Wind¶

Round 2 hierarchical compression | Synthesised from 45 batch summaries (~1,200 unique papers) | 2026-04-17

Domain at a glance¶

mindmap
  root((D1 · Geotechnical<br/>1 200 papers))
    Established knowledge
      V-H-M failure envelopes
        Butterfield 1994
        Gourvenec 2007-08
        Bransby 1998-2009
        Skau 2018
        Jin 2025
      SSI controls OWT frequency
        Arany 2015-16
        Jalbi-Bhattacharya 2018
        Stuyts 2022-23
        Zaaijer 2006
      Suction caissons viable
        Bothkennar clay
        Luce Bay sand
        Borkum Riffgrund 1
        OxCaisson Suryasentana 2018
      Scour degrades foundation
        Prendergast 2013-17
        Li 2020-21
        Qi-Gao 2014-19
        DNV 1.3D rule
      PISA > API p-y
        Beuckelaers 2017
        McAdam 2023
        Haiderali 2023
      Cyclic stiffens sand
        LeBlanc 2010
        Cox 2014
        Abadie HARM 2015
      Centrifuge gold standard
        KAIST Cambridge UWA
        ETH Oxford Aalborg Delft
    Active frontiers
      Large-diameter monopiles
      Scour-cyclic coupling
      Machine-learning SSI
      Asymmetric scour
      Silty-sand mechanics
    Methods toolkit
      VHM envelope FEA
      Winkler p-y
      Centrifuge modelling
      OptumGX limit analysis
      OpenSees dynamics
    Relevance to PhD
      J1 centrifuge tripod
      J2 3D FE Winkler
      J3 saturation backfill
      J5 probabilistic capacity
      J11 Vesic generalisation

Key anchor citations on a timeline¶

%%{init: {"theme":"base","themeVariables":{"fontSize":"16px","fontFamily":"Inter, system-ui, -apple-system, sans-serif","primaryTextColor":"#1a1a1a","lineColor":"#666"},"flowchart":{"nodeSpacing":18,"rankSpacing":22,"padding":14,"useMaxWidth":false}}}%%
flowchart TB
    T1994["<b>1994 ·</b> Butterfield &amp; Gottardi — VHM envelope for circular footings"]:::y
    T1998["<b>1998 ·</b> Bransby &amp; Randolph — skirted foundation envelopes"]:::y
    T2005["<b>2005 ·</b> Houlsby — Bothkennar suction caisson (clay)"]:::y
    T2006["<b>2006 ·</b> Zaaijer — mudline stiffness matrix"]:::y
    T2007["<b>2007 ·</b> Gourvenec — rectangular + embedded VHM"]:::y
    T2010["<b>2010 ·</b> LeBlanc — cyclic stiffening in sand (8k-60k cycles)"]:::y
    T2014["<b>2014 ·</b> Cox — suction caisson cyclic stiffening (12k cycles)"]:::y
    T2015["<b>2015 ·</b> Arany — three-spring closed-form OWT frequency"]:::y
    T2016["<b>2016 ·</b> Arany — &lt;3.5% error across 10 European wind farms"]:::y
    T2017a["<b>2017 ·</b> Beuckelaers — PISA framework launch"]:::y
    T2017b["<b>2017 ·</b> Shonberg — Borkum Riffgrund 1 SBJ monitoring"]:::y
    T2018["<b>2018 ·</b> Suryasentana — OxCaisson Winkler surrogates"]:::y
    T2022["<b>2022 ·</b> Stuyts — monitored f is 5–15% above design"]:::y
    T2023["<b>2023 ·</b> McAdam — PISA beats API on field data"]:::y
    T2025["<b>2025 ·</b> Jin — scour depth dominates capacity loss"]:::y

    T1994 --> T1998 --> T2005 --> T2006 --> T2007 --> T2010
    T2010 --> T2014 --> T2015 --> T2016 --> T2017a --> T2017b
    T2017b --> T2018 --> T2022 --> T2023 --> T2025

    classDef y fill:#e3f2fd,stroke:#1565c0,stroke-width:2px,color:#0d47a1,text-align:left

1. ESTABLISHED KNOWLEDGE¶

1.1 Combined V-H-M loading governs offshore foundation design¶

The failure envelope framework -- not classical superposition -- is the accepted method for assessing multi-directional capacity. Butterfield & Gottardi (1994) established the experimental VHM envelope for circular footings on sand (cigar-shaped, elliptic cross-sections with parabolic meridians). Butterfield et al. (1997) standardised VHM sign conventions adopted universally. Gourvenec (2007, 2008) extended the framework to rectangular footings, embedment depth factors, and heterogeneous undrained clay via 3D FEM. Bransby & Randolph (1998) and Bransby & Yun (2007, 2009) developed failure envelopes for skirted foundations. Fu (2017, UWA) showed that preloading with consolidation substantially expands VHM capacity. Vulpe (2013-2017) generalised to skirted spudcans and consolidated undrained conditions using critical-state FEM. Skau et al. (2018, Geotechnique) demonstrated that the normalised shape of VHM failure envelopes is approximately invariant with number of equivalent cycles in clay. Jin et al. (2025) showed that scour depth dominates bearing capacity loss over scour width/angle, but the failure envelope shape itself remains unchanged.

1.2 Soil-structure interaction controls OWT natural frequency¶

Every study in the corpus confirms that fixed-base assumptions overestimate natural frequency. The three-spring model (lateral KL, rotational KR, cross-coupling KLR) is the standard for monopile OWTs (Arany et al. 2015, 2016; Jalbi & Bhattacharya 2018). Arany et al. (2016) achieved <3.5% error across 10 European wind farms using a closed-form Euler-Bernoulli beam on three-spring foundations. Stuyts et al. (2022, 2023) documented that monitored natural frequencies are systematically 5-15% higher than design predictions, indicating conservative stiffness assumptions in current practice. Zaaijer (2006) showed that a mudline stiffness matrix suffices for ~4% accuracy.

1.3 Suction caissons are viable OWT foundations¶

Field trials at Bothkennar (clay, Houlsby et al. 2005) and Luce Bay (sand, Houlsby et al. 2006) provided the first prototype-scale validation. Borkum Riffgrund 1 delivered the world's first full-scale suction bucket jacket (SBJ) with operational monitoring (Shonberg et al. 2017). OWA/Carbon Trust (2019) published comprehensive design guidelines covering ULS, SLS, FLS, and ALS limit states. Efthymiou & Gazetas (2018, 2022) showed that for L/R > 1, the cylindrical sidewall shell alone provides nearly the full stiffness and lateral capacity. Suryasentana (2018, DPhil) developed the OxCaisson family of Winkler-based surrogate models (LE/NLE/LEPP) that deliver 3D-FEM-comparable results at orders-of-magnitude lower cost. Typical L/D is ~1:1 in sand, ~2:1 in clay.

1.4 Scour degrades foundation performance¶

Unanimous across monopiles, suction buckets, and bridge piers: scour reduces stiffness, lowers natural frequency, and degrades lateral/moment capacity (Prendergast 2013-2017; Li et al. 2020, 2021; Ma 2018; Tseng 2017, 2018; Qi & Gao 2014-2019). Global scour is more detrimental than local scour for the same nominal depth because it removes overburden stress across a wider area (Li 2020; Kariyawasam 2020; Ciancimino 2022). The DNV 1.3D scour depth rule is a standard design assumption but is context-dependent and may be non-conservative for complex geometries. Robin Rigg wind farm lost two monopiles to unexpected scour after 6 years of operation (Vicente/Carlos 2023).

1.5 API p-y curves are inadequate for large-diameter monopiles¶

The PISA framework (Beuckelaers 2017; Byrne et al.) provides substantially better agreement with monitored data than API/DNVGL p-y methods (Jurado 2022; Kheffache 2024; McAdam 2023). Haiderali et al. (2023) showed via centrifuge-calibrated 3D FEA that the p-y method is inaccurate for XL/XXL monopiles in clay. Choo & Kim (2015) showed API overestimates initial stiffness for large-diameter monopiles in sand. Mozaffari et al. (2024) developed an extended Winkler model matching PISA predictions for large-diameter piles.

1.6 Cyclic loading stiffens sand, may soften clay¶

LeBlanc (2010) demonstrated that monopile stiffness in sand increases with number of cycles (8,000-60,000 cycles), contradicting API's degradation assumption. Cox et al. (2014) showed logarithmic stiffness gain for suction caissons in dense sand over 12,000 cycles. Barari et al. (2021) documented "self-healing" rotation mechanisms in tripod suction caissons in silty sand. In contrast, Lombardi (2013) found frequency changes in clay depend on shear strain level. Abadie (2015, Oxford) developed the HARM model within hyperplasticity capturing ratcheting in cohesionless soils following extended Masing rules.

1.7 Centrifuge testing is the gold standard¶

Across all 45 summaries, centrifuge testing is universally treated as the primary validation tool for offshore geotechnical problems, used at KAIST, Cambridge, UWA, ETH, Oxford, Aalborg, and Delft. Kim et al. (2014) tested tripod buckets at 100g; Cox et al. (2014) tested caissons at 1:200; Garala (2022) validated kinematic pile bending; Espanol-Espinel (2024) validated composite monopile+outrigger systems in liquefiable soils.

2. ACTIVE FRONTIERS (2023-2025)¶

2.1 Macro-element and data-driven foundation surrogates¶

The Skau/NGI body of work (2015-2019) developed multi-surface plasticity macro-elements calibrated against 3D FEM for buckets and monopiles. Jin et al. (2019, 2025) extended hypoplastic macro-elements to include scour as a single parameter. Lopez et al. (2025) proposed the first ANN-based macro-element for suction buckets in 3D VHM space. Orakci et al. (2023) combined PISA p-y springs with Arany's simplified method for rapid natural frequency estimation, validated against OpenSees.

2.2 Physics-informed ML for geotechnics and SHM¶

Cross et al. (2021) demonstrated physics-informed Gaussian process regression for SHM. Liu et al. (2023) developed a physics-informed neural network replacing SVD-based modal analysis for real-time offshore monitoring. Shen et al. (2023, Nature Reviews) advocated differentiable modelling embedding physical equations within neural network training loops. Smith et al. (2023) used hierarchical Bayesian models for population-level scour detection across a wind farm.

2.3 Vibration-based scour monitoring advancing toward quantification¶

Weil et al. (2023) proposed digital twin + automated OMA for quantifiable (not just detectable) scour assessment. Zou et al. (2022) achieved the first in-situ validation of GPLFM for below-mudline strain estimation at Westermeerwind Park. Kim et al. (2026, submitted) developed a double-filter (RANSAC+CUSUM) framework that reduces EOV-induced frequency scatter by 70% and detects scour at 0.39D with 95% probability and zero false alarms, using 32 months of field data from a 4.2 MW tripod. Kim et al. (2025) measured max 5.3% frequency reduction at scour depth = 0.6D in centrifuge tests of tripod suction buckets.

2.4 Probabilistic digital twins for OWT¶

Bull et al. (2025) argued for ensembles of probabilistic digital twins rather than single deterministic models. Branlard et al. (2024) validated a full-scale floating turbine digital twin at 10-15% error on fatigue DELs using Kalman filter + aerodynamic estimator on TetraSpar data. Tian, Wang & Phoon (2024) developed physics-informed sparse dictionary learning for real-time geotechnical data fusion.

2.5 Silent/screw piling for offshore¶

Cerfontaine et al. (2021, 2023) demonstrated screw pile installation by rotary jacking at low reaction force via DEM + centrifuge. Davidson et al. (2020) showed screw piles offer quieter installation and enhanced tensile capacity versus driven piles. Wang et al. (2023) showed over-flighting (AR < 1) enhances cyclic uplift performance. Regulatory acceptance and long-term field validation remain outstanding.

3. CONTESTED CLAIMS¶

3.1 Associated vs. non-associated flow in plasticity models¶

Houlsby (1981) rigorously derived non-associated flow from thermomechanics. The Oxford plasticity models (Martin 1994; Cassidy 1999) often use associated or nearly-associated flow for convenience. The error introduced by this simplification is debated and varies by load path and soil type.

3.2 Local vs. general scour treatment in p-y methods¶

Qi et al. (2016) showed local scour leaves overconsolidation effects that make soil stiffer than mudline-lowering (general scour) would predict. Ciancimino (2022) showed local scour reduces Mult by 38% vs. 48% for general scour -- fundamentally different failure mechanisms. No standardised code provision exists.

3.3 Drained vs. undrained bucket behaviour under cyclic loading¶

Nielsen (2017) showed two-way loading produces the largest rotation under partly drained conditions (contrary to drained findings where one-way dominates). Choo et al. (2021) showed 1.5x capacity increase at fast loading rates due to dilation-induced pore pressure. Shonberg et al. (2017) identified drainage boundaries from field pore-pressure data. The transition is unresolved.

3.4 Macro-element vs. p-y vs. continuum FEM for design¶

Sorum et al. (2022) showed macro-element and p-y models yield materially different fatigue damage estimates. Skau (2019) showed lid flexibility matters for jacket caissons (not captured by most macro-elements). Industry retains p-y due to simplicity and regulatory acceptance; research advocates macro-elements. The transition path into design codes is contested.

3.5 Rigid vs. deformable soil plug in skirted foundations¶

Bransby & Yun (2009) showed the assumption of rigid soil between skirts does not always hold, contradicting earlier work. Vulpe (2015) showed deformable plug significantly affects failure envelope shape. The "correct" assumption is site- and geometry-dependent.

3.6 Frequency vs. mode shape as scour indicator¶

Kariyawasam (2019a) found that field frequency variability can exceed scour-induced shifts, questioning frequency-only methods. Khan (2021) and Malekjafarian (2020) advocate mode-shape-based indicators (MNMS, MAC) as more robust. Jawalageri (2022) showed second mode is more sensitive but requires more sensors.

4. VERIFIED GAPS (confirmed by multiple independent sources)¶

Combined scour + cyclic loading on OWT foundations. Scour studies ignore cycling; cyclic studies ignore scour. No experimental or numerical study couples progressive scour with millions of load cycles. (Confirmed by: batch01-agent1, batch01-agent2, batch02-agent2, batch05-agent2, batch05-agent5, batch06-agent3.)
Scour effects on suction bucket/caisson foundations. Research is overwhelmingly concentrated on monopiles and bridge piers. Only Chen (2018), Cheng (2024), Kim (2025), and Jin (2025) address bucket-specific scour. No centrifuge study combines in-flight scour with cyclic loading for multi-bucket foundations. (Confirmed by: batch05-agent2, batch05-agent5, batch06-agent3.)
Long-term field validation of frequency-based scour monitoring for OWT. Concept is well-established in labs and bridge field trials (Prendergast, Kariyawasam). No fully operational field system for OWT has been published. Kim et al. (2026, submitted) represents the closest with 32 months of tripod data. (Confirmed by: batch04-agent1, batch05-agent5, batch06-agent5.)
Prototype-scale validation of centrifuge-derived design methods. Universally called for since the 1990s (Bell 1991, Martin 1994) and still lacking (batch01-agent1, batch02-agent5, batch03-agent2).
Cyclic macro-elements for sand. Skau's macro-element library is developed primarily for clay (Drammen clay cyclic contours). Extension to sand with drained/undrained transitions is not at the same maturity. (Confirmed by: batch06-agent2.)
Cohesive soil scour. Most scour studies use sand. Harris (2023) explicitly states large uncertainty in cohesive marine soils. No widely accepted predictive method exists. (Confirmed by: batch05-agent4, batch06-agent4.)
Multi-hazard loading (scour + seismic + fatigue). Individual hazards are treated separately. Ngo (2022) and Jia (2024) are among the very few combining two. A unified probabilistic multi-hazard framework does not exist. (Confirmed by: batch02-agent2, batch03-agent3, batch04-agent1.)
Digital twin for soil-foundation systems. Branlard (2024) covers the superstructure; Al-Subaihawi (2024) introduces RTHS for soil-pile interaction. No integrated digital twin spans the full aero-hydro-geotech chain with validated soil modelling. (Confirmed by: batch04-agent3.)

5. QUANTITATIVE BENCHMARKS¶

Foundation dimensions and capacity¶

Parameter	Value	Source
Monopile market share for OWT	77-80%	Abdelhak 2024; Weijtjens 2017
Foundation cost share of OWT	15-40% of capital	Houlsby & Byrne 2000; Tran 2017
Tripod bucket dimensions (KAIST)	D=6.5 m, L=8.0 m, S=26.9 m at 100g	Kim et al. 2014
Hybrid bucket capacity gain	1.91x (V), 1.82x (combined) over SBF	JH Kim et al. 2020
Tripod optimal spacing for H-capacity	S/D = 1.5-3.5 (for L/D = 0.5-1)	Tran et al. 2017
Suction caisson critical uplift displacement	~0.02D (mono), ~0.01D (group) in soft clay	Zhu et al. 2019
Cyclic preload gain over monotonic (clay)	+50% bearing capacity	Vulpe & White 2014

Scour¶

Parameter	Value	Source
DNV design scour depth	1.3D (current-only)	DNV-RP-0618
Max scour depth (wave+current, live-bed)	S/D up to 2.0	Whitehouse 2004
Monopile frequency drop per 1D scour	5-15% (soil/L/D dependent)	Li 2020; Jawalageri 2022
Tripod bucket frequency drop at 0.6D scour	5.3% max	Kim et al. 2025
Bridge frequency change for 30% embedment loss	Up to 40%	Kariyawasam 2020
Scour-induced fatigue life reduction (1.3D)	~24%	Cao 2024
TMD fatigue life increase (1% mass ratio)	~65% at 1.3D scour	Cao 2024
Lateral displacement reduction from scour protection	Up to 41% (2D width, 30 kPa)	Q. Li et al. 2024

SSI and stiffness¶

Parameter	Value	Source
Design underestimation of natural frequency	5-15% (consistently)	Stuyts 2022/2023
Arany closed-form fn error	<3.5% across 10 wind farms	Arany et al. 2016
Ryu closed-form tripod fn error	<1% vs field	Ryu et al. 2020
Foundation model choice fatigue variation	Up to 22% (monopile)	Aasen et al. 2017
Foundation model choice fatigue variation	Up to 180% (misaligned wind-wave)	Katsikogiannis 2019
Low-frequency fatigue dynamics contribution	Up to 65% of total damage	Sadeghi et al. 2023
Soil damping ratio range for monopile	0.17-1.3% of critical	Rezaei et al. 2018
SSI demand reduction (inelastic soil)	>50% superstructure demand	Homaei 2020
G/G0 degradation prediction accuracy	Factor 1.13 (1 SD, 3860 points)	Oztoprak & Bolton 2013

Key datasets and reference systems¶

Dataset	Description	Source
Butterfield/Gottardi (1994, 1999)	Foundational VHM envelope for circular footings on sand	Oxford/Padova
Gazetas (1991)	Canonical impedance functions (spring-dashpot) for foundations	NTUA
PISA field tests	Large-scale at Dunkirk (sand) and Cowden (clay)	Oxford/Imperial
Bothkennar/Luce Bay field trials	Suction caisson in clay/sand	Oxford (Houlsby 2005/2006)
Borkum Riffgrund 1 SBJ monitoring	First full-scale suction bucket jacket	DONG/SPT
OxCaisson (Suryasentana 2018-2023)	Winkler surrogate hierarchy for suction caissons	Oxford
JCSS PMC Sec 3.7	Soil property distributions, scales of fluctuation, CoV	JCSS 2006
Ching CLAY-Cc/6/6203	6203 clay compressibility records from 429 studies	NUS
NREL 5/10/15 MW reference turbines	Standard turbines for OWT research	NREL/DTU/IEA
Robin Rigg failure case	2 monopiles decommissioned after scour	Carlos/Vicente 2023
Cox centrifuge caisson tests	12,000 cycles at 1:200 in dense sand	Oxford

Synthesised from batch01_agent1 through batch09_agent5 (45 files, ~1200 unique papers). This file serves as Round 2 input for the master knowledge map.