PSI - Issue 78

Emanuele Brunesi et al. / Procedia Structural Integrity 78 (2026) 161–168

162

exhaustive information on both induced seismicity as well as proposals for risk-informed decisions and mitigation actions can be found elsewhere (see e.g. Bommer et al. 2015), this paper presents the main findings of an experimental activity that was carried out at the Eucentre laboratory (Pavia, Italy), as a part of a broad research programme on the seismic vulnerability of existing Dutch buildings. Indeed, given the lack or very limited availability of data related to the seismic behaviour of construction typologies specific to Dutch building practice, a comprehensive testing campaign was undertaken within the framework of the research programme for hazard and risk of induced seismicity in Groningen sponsored by the Nederlandse Aardolie Maatschappij BV (NAM), with a view to study the performance of structural components and systems representative of the Groningen reinforced concrete (RC) building stock. These structures largely consist of what are herein termed wall-slab-wall structures and can be found in both cast-in-place (tunnel construction or not) as well as precast configurations, both of which were tested experimentally in pseudo-static cyclic fashion and/or dynamically by shake table. Although the full-scale building specimens with which this paper is concerned were meant to resemble the main features of typical terraced houses, and hence to follow structural layouts and construction details that are typical in the Netherlands for this residential typology, structural systems of this kind are relatively widespread and can be found in both countries exposed to potential large tectonic earthquakes such as Turkey, Iran, Armenia, Chile, New Zealand, US, as well as nations subjected to small or medium-size natural/anthropogenic seismic events such as Germany, UK, Australia, in addition to the Netherlands – see e.g. Hawkins et al. (1994), Glass (2000), Wilson et al. (2008). Thus, the experimental test results obtained and collected in this paper, along with previous separate contributions (Brunesi et al. 2019a; 2018a; 2018b; Brunesi and Nascimbene 2017) are not only Groningen specific and could be helpful not only for the Groningen case study but could also serve as a more general reference for similar structures around the world. The objectives of the tests presented herein, and more generally of the entire testing campaign, can therefore be summarised as follows: i) to investigate the seismic response, identifying main deficiencies, of the non seismically designed structural systems that the test specimens are intended to represent; ii) to evaluate the capacity of the structures when subjected to cyclic pseudo-static loading or to a dynamic input motion consistent with local seismic hazard or higher than that, through the scaling of the input motions to assess structural capacity up to the near collapse state; and iii) to identify and quantify dominant response mechanisms and damage modes, providing also information for potential numerical modelling efforts. In relation to this, it is noted that the test results have been employed by Crowley et al. (2017; 2019) in the development of fragility functions that were then employed in the probabilistic seismic risk analysis for the region (van Elk et al. 2019). Lastly, it is worth mentioning that the experimental test results of the three full-scale buildings presented herein, namely a one-storey two-bay cast-in-place one tested in pseudo-static cyclic fashion and two two-storey one-bay precast ones tested in that manner as well as on the large-size uniaxial shake table of Eucentre, are openly made available at the Experiments platform of the Built Environment Data initiative, along with auxiliary connection testing and friction testing (Brunesi et al. 2019b; 2020). The identified structural deficiencies (e.g. connection behavior for the precast configuration and rocking of walls for the cast-in-place system, together with sliding/slipping of walls along starter rebars once complete bond at the concrete-steel interface is lost) could potentially be also employed towards the definition of performance criteria and the revision of structural design codes in the Netherlands (as well as in the design of structural upgrading interventions and in the deployment of seismic retrofitting campaigns). 2. Construction system and characteristics of full-scale test specimens As can be gathered from Figure 1, reinforced concrete terraced houses in Groningen are generally one- or two storey multi-unit buildings, whose structural layout in precast configuration consists of precast hollow core floor slabs, precast party/gable walls and precast load-bearing walls – in both longitudinal and transverse directions – that are connected together by means of mortar joints, felts and mechanical steel anchors, with the latter being used for the wall-to-wall joints only. Such connections consist not only of two-way anchorages (for contiguous lateral panels) but also of three-way links, given that contiguous panels are often connected to an orthogonally placed stability/shear wall, which is a specific feature the two nominally identical full-scale precast building specimens reproduced (Figure 2). It is particularly noteworthy the fact that no starter rebars are present at the base of the first-storey walls, with the same applying also to the second-floor lateral and stability walls, where no rebars connect the first- with the second storey walls, through the floor (see Figure 3). Because of the above (i.e. absence of steel rebars protruding into the