Issue 78

S. Majer et alii, Fracture and Structural Integrity, 78 (2026) 504-518; DOI: 10.3221/IGF-ESIS.78.29

Figure 12: Pavement design of the intermodal railway terminal utilizing a reinforced concrete slab system.

In the case of logistics yards for offshore wind farms, nacelle storage areas, and installation terminals, unbound aggregate pavements reinforced with geosynthetics are commonly utilized. These solutions consist of thick layers of crushed aggregate stabilized with geogrids. Operational experience indicates that such pavements perform satisfactorily under heavy static loads generated by stored wind turbine components. An additional advantage of aggregate pavements is their ease of maintenance; local deformations can be rapidly corrected by grading and replenishing the aggregate material without interrupting terminal operations. This characteristic is particularly vital for offshore wind farm terminals, where operational requirements and storage space layouts may change dynamically throughout the facility's lifespan. As an example, the settlement analysis results for the aggregate pavement at the offshore installation terminal in Ś winouj ś cie are presented. The heterogeneous subsoil comprised uncontrolled anthropogenic fills, sandy deposits locally interbedded with peat and organic silt/gyttja, and shallow groundwater. The high and long-duration contact pressures generated by stored monopiles required an elasto-plastic finite-element assessment. The analysis was performed in GEO5 (FEM – Stress module) under two-dimensional plane-strain conditions. Fig. 13 shows the calculated settlement field for the governing monopile storage case. The numerical domain was 108.0 m wide and 20.0 m deep. The pavement surface was represented by an approximately 0.69 m thick equivalent structural layer. The laterally varying soil boundaries reproduced the site investigation profile; within the loaded central part of the model the near-surface fill and sand package was underlain by discontinuous peat and organic silt/gyttja layers and, at greater depth, by medium sands. The vertical side boundaries were restrained horizontally (u ₓ = 0) and remained free vertically, while the base was fixed in both directions (u ₓ = u_z = 0). The ground surface was traction free outside the loaded area. The domain was discretized using higher-order multi-node triangular elements with a target element size of 0.50 m in the loaded zone and along material interfaces; automatic smoothing was applied outside this zone. A mesh-refinement check using characteristic sizes of 0.75, 0.50 and 0.35 m changed the maximum settlement by less than 2.5% between the two finest meshes, and the 0.50 m mesh was therefore adopted for the production analyses. Groundwater was represented by a horizontal phreatic surface 0.17 m below the reference ground level. The consolidation calculation was formulated in effective stresses and included pore pressure dissipation; saturated unit weights and directional coefficients of permeability were assigned to the soil layers. The load was applied as a uniformly distributed vertical pressure q = 150 kPa over a 30.0 m wide central strip (x = 45.0–75.0 m), representing the equivalent bearing area of the monopile storage supports in the plane-strain model. The load was ramped within one construction stage in four increments (25, 50, 75 and 100%) and was subsequently maintained during the 356-day consolidation period. The soil response was represented by a Drucker–Prager yield surface calibrated to the effective Mohr–Coulomb strength parameters on the triaxial- extension meridian. The adopted yield function was [14]:

515