Skip to content

J1 — Scour, Natural Frequency, and Sensitivity for Tripod Suction Bucket Foundations in Sand

Journal: Ocean Engineering (2025), article 123084. Status: Published (accepted 2025-10-04). Authors: Kim, K.-S.; Oh, S.-W.; Kim, B.-S.; Kim, S.-R. DOI: 10.1016/j.oceaneng.2025.123084

Read the author-accepted manuscript (PDF, 1.0 MB) Journal link

Author's accepted manuscript

This page is the author's accepted manuscript (AAM) of a paper published in Ocean Engineering, 342 (2025), article 123084. The text below is the post-peer-review revision; the publisher's typeset version (the version of record) is authoritative.

Version of record: https://doi.org/10.1016/j.oceaneng.2025.123084

Shared under CC BY-NC-ND 4.0, in accordance with Elsevier's author-sharing policy.

(No editable .qmd source is retained for this paper; the sections below are the portfolio outline.)

Full title

Investigating Scour Impacts on Natural Frequency Changes and Sensitivity in Offshore Wind Turbine with Tripod Suction Bucket Foundations in Sand.

One-sentence headline

The first systematic centrifuge dataset on scour-induced natural-frequency sensitivity for a tripod suction bucket (TSB) foundation in dry sand, establishing the CPT-bridging procedure and the power-law scour-frequency relationship that every later paper in the portfolio builds on.

Context

Offshore wind foundations are moving into deeper water and stronger current regimes, which makes scour the dominant long-term geotechnical hazard. The industry's design tools were built for slender monopiles with aspect ratios \(L/D > 10\); tripod suction buckets operate at \(L/D \approx 1\) and redistribute overturning load through three discrete footings rather than through embedment along a single shaft. That difference is not a detail — it changes which mechanisms govern scour sensitivity and therefore which monitoring indicators are reliable. Before this paper, no centrifuge programme had measured the scour-frequency relationship for a TSB at all, and the numerical tools used to estimate it were calibrated on monopiles.

The paper matters for the portfolio because it fixes the experimental baseline on which every later paper either extends (J3 to saturated conditions) or compares against (J2's numerical model, B's cross-domain features).

Research question

How does local scour affect the natural frequency and dynamic sensitivity of tripod suction bucket foundations in sand, and how can CPT-based scaling allow centrifuge measurements to be compared with full-scale prototype predictions without introducing extra calibration free parameters?

Approach

The paper runs a 70 g centrifuge programme on a 1:70 scale model of the Gunsan 4.2 MW tripod with three 8 m × 9.3 m suction buckets. Three dry-sand series are tested: dense No. 7 sand, loose No. 7 sand, and a sand–silt mix. Each series runs four progressively deeper scour stages (\(S/D = 0\) to ~0.6). Dynamic excitation uses impulse, sinusoidal, and square-wave sequences; the first natural frequency is extracted from the accelerometer record by operational modal analysis. In-flight cone penetration testing quantifies the soil state for every flight, and the CPT profile is the normalising quantity that translates model-scale measurements to prototype-scale predictions without an extra stiffness fit.

The power-law form \(|\Delta f / f_0| = a \cdot (S/D)^b\) is extracted by least-squares fit across all dense-sand scour stages, giving the first quantitative scour-frequency curve for a TSB in sand. The CPT-bridging procedure — normalising shear-strength profiles between centrifuge and field via cone tip resistance — is the methodological legacy that subsequent portfolio papers depend on.

Gap the paper closes

  • Defensive. No centrifuge evidence existed for the scour-frequency sensitivity of a tripod suction bucket foundation in sand.
  • Offensive. The monopile-calibrated empirical relationships in every design code were being quietly applied to TSB foundations with no experimental test of their validity — a thirty-year extrapolation beyond the tested \(L/D\) regime.
  • Constructive. Supplies the dry-sand baseline that J3 extends to saturation, the experimental reference that J2 validates against, and the feature-extraction pipeline reused by Paper B.

Key literature anchors

  • Prendergast et al. (2015) — laboratory frequency–scour demonstration on piles.
  • Houlsby (2017) — suction caisson mechanics synthesis (lid bearing, skirt friction, passive wedge).
  • Kallehave et al. (2015) — field survey of ~400 monopile-supported turbines showing measured frequency exceeds design predictions by up to 20 %.
  • Madabhushi (2014), Bhattacharya (2013) — centrifuge modelling fundamentals for offshore foundation problems.
  • Jalbi & Bhattacharya (2018, 2020) — closed-form expressions for jacket and tripod natural frequency; multi-footing frequency sensitivity is about 5 × lower than monopile.

Headline findings

  1. Frequency decline is monotonic with scour depth and follows a power law with soil-dependent coefficients.
  2. Dense and loose sand produce different scour-frequency slopes; the loose-sand slope is substantially steeper.
  3. CPT-normalised comparison between centrifuge and prototype scale is internally consistent, establishing the bridging procedure for later papers.
  4. Even at \(S/D = 0.5\)–\(0.6\) the frequency change is modest (5 %), confirming that TSB foundations are less scour-sensitive than monopiles at operational depths.

Limitations

  • Dry sand only — the dominant field condition is saturated, addressed in J3.
  • Single geometry (Gunsan prototype) — generalisation to other bucket aspect ratios is outside the dry-sand dataset.
  • Local scour geometry — global scour not tested.
  • The centrifuge protocol excavates at 1 g and spins up, which produces a mild OCR in the near-scour zone. The companion paper J3 flags this as a 15–25 % stiffness bias and reframes the centrifuge result as a lower bound on scour sensitivity.

Portfolio flow

  • Consumes: nothing upstream; this is the experimental anchor of the portfolio.
  • Produces: dry-sand power-law coefficients, CPT-bridging procedure, test-matrix template, data-processing pipeline reused by J3 and B.

Status

Published and closed. Used as the experimental baseline across the rest of the portfolio; full working tree retained at the original location for cross-citation during J3 revision.