Soil–structure interaction (SSI) — the modification of a building's seismic response by the flexibility and damping of its supporting soil and foundation — is traditionally assumed to be beneficial, and seismic design provisions often allow it to be neglected. A large probabilistic study (Khosravikia, Mahsuli and Ghannad, 2018) showed that this assumption fails for loss risk: on very soft soils SSI is extremely likely to reduce seismic monetary losses, but on moderately soft soils there is a considerable probability, up to about 0.4, that SSI increases them. Those results are outcome statistics over the studied building population; the conditions that flip the sign of the SSI effect are not known in general. The open problem is to establish general, validated criteria that determine a priori whether SSI is beneficial, detrimental, or uninfluential for the probabilistic seismic monetary loss of a building, as a function of soil stiffness, structural system, story count, foundation type and size, and ground-motion characteristics, with the intended application of translating them into engineering design guidance stating when neglecting SSI is unconservative.

Background

Soil–structure interaction (SSI) denotes the dynamic coupling between a building, its foundation, and the supporting soil: soil flexibility lengthens the building's fundamental period (the natural sway period of the structure), while foundation motion dissipates energy through radiation and hysteretic damping in the soil. Because modeling this coupling directly is expensive, seismic design practice traditionally treats buildings as fixed at the base, and provisions such as ASCE 7 and the NEHRP provisions permit SSI to be neglected in many cases. This practice rests on the long-standing belief, rooted in elastic analysis, that SSI is beneficial: period lengthening shifts the structure away from the peak of the response spectrum (the curve of peak structural response versus period used to characterize ground-motion severity), while foundation damping removes energy. The sign question is nevertheless old: Mylonakis and Gazetas posed it explicitly in 2000 — "seismic soil–structure interaction: beneficial or detrimental?" — and at the level of elastic response the answer is known to depend on the interplay between the lengthened system period and the spectral shape of the ground motion.

Modern performance-based earthquake engineering measures earthquake consequences instead by probabilistic seismic loss: the probability distribution of monetary loss, obtained by combining ground-motion hazard with fragility curves (probabilities of exceeding damage states as a function of a response measure such as interstory drift, the relative sideways displacement between adjacent floors) and repair-cost models. Khosravikia, Mahsuli and Ghannad (2018) placed SSI inside such a framework: they coupled probabilistic response, damage, and repair-cost models with SSI models for a wide range of buildings — different structural systems, numbers of stories, and foundation sizes on various soil types — and a suite of nearly 7000 accelerograms (recorded ground-motion time histories) on soil, with uncertainty in both the ground motions and the soil–structure-system properties. Their risk analysis reports the probability that SSI is beneficial, detrimental, or uninfluential on seismic losses. Their central finding is that the traditional belief holds only for very soft soils, where SSI is extremely likely to reduce losses; on moderately soft soils they observed a considerable probability, up to 0.4, that SSI increases seismic losses. In the group's companion framework, the considered soils are classified by ASCE 7 site classes C, D and E, representing moderate to very soft soils. The same research group had earlier evaluated the SSI provisions of design codes probabilistically (NEHRP 2015 provisions; ASCE-7 code provisions), and later extended the probabilistic treatment to strength demands of nonlinear shear buildings (Abtahi, Mahsuli and Ghannad, 2020) — that is, to structural demand quantities rather than to monetary loss.

The 2018 study thus established that the sign of SSI's effect on seismic loss risk is not uniformly beneficial, but it delivered outcome statistics over the studied building population. What does not yet exist — and what subsequent literature has not supplied — is a general set of conditions, in terms of soil stiffness (for instance soil class or shear-wave velocity, the speed of seismic shear waves in the soil), structural system, story count, foundation type and size, and the governing ground-motion characteristics, that predicts a priori on which side of the sign boundary a given building lies, in a form usable as engineering design guidance. Later work confirms the phenomenon and illuminates the mechanism (the period-shift mechanism, quantified response-amplification probabilities, case studies with large loss increases) but stops short of general loss-level sign criteria.

Problem Statement

Establish general criteria for the sign of soil–structure interaction (SSI) effects on the seismic monetary-loss risk of buildings. Specifically: determine and validate conditions that predict, before a project-specific probabilistic analysis is run, whether including SSI is beneficial, detrimental, or uninfluential for the probabilistic seismic monetary loss of a building — that is, in the three-category classification of the source study (Khosravikia, Mahsuli and Ghannad 2018), where the probability that SSI increases the seismic monetary loss is low (beneficial), high (detrimental), or within a neutral band (uninfluential), each relative to the conventional fixed-base comparison — as a function of soil stiffness (for example soil class or shear-wave velocity), structural system, number of stories, foundation type and size, and the governing ground-motion and hazard characteristics. Here the listed factors delimit the scope of building–soil configurations over which the criterion must be general rather than a conjunctive requirement to parameterize every axis separately; the criterion is required to capture at least the dependence on soil stiffness and on the structural period (through the structural system and story count). Following the source group's convention, the sign object is the probability, within the probabilistic loss framework, that including SSI increases the seismic monetary loss relative to the fixed-base comparison; the three categories are regions of this probability, and any neutrality band delimiting the 'uninfluential' category must be explicitly specified and justified (the group's companion studies interpret probabilities near 0.5 — for example a 0.4–0.6 band — as indicating nearly the same median performance as the fixed-base design, although those particular thresholds belong to their code-provision metric rather than to the loss study itself, which does not fix such a band). The criteria must be formulated and validated within a probabilistic loss assessment framework that accounts for ground-motion record-to-record variability and uncertainty in the soil–structure system. Following the source study's findings, the criteria should recover the reported direction of the effect as a qualitative consistency check: SSI extremely likely to reduce losses on very soft soils, and a considerable probability, up to about 0.4, of increasing losses on moderately soft soils. Those figures are results for that study's building population, design-basis hazard and loss models rather than universal benchmarks, so the check is on the sign and soil-class ordering, not on an exact numerical reproduction. Translating the criteria into engineering design guidance stating for which building–soil–foundation configurations neglecting SSI in design is unconservative is the intended application of the results.

References

  1. Khosravikia, F., Mahsuli, M., Ghannad, M. A. (2018). The effect of soil–structure interaction on the seismic risk to buildings. Bulletin of Earthquake Engineering 16(9):3653–3673. DOI: 10.1007/s10518-018-0314-z. https://doi.org/10.1007/s10518-018-0314-z (source study)
  2. Mylonakis, G., Gazetas, G. (2000). Seismic soil–structure interaction: beneficial or detrimental? Journal of Earthquake Engineering 4(3):277–301. DOI: 10.1080/13632460009350372. https://doi.org/10.1080/13632460009350372
  3. Khosravikia, F., Mahsuli, M., Ghannad, M. A. (2017). Probabilistic Evaluation of 2015 NEHRP Soil-Structure Interaction Provisions. Journal of Engineering Mechanics 143(9):04017065. DOI: 10.1061/(ASCE)EM.1943-7889.0001274. https://doi.org/10.1061/(ASCE)EM.1943-7889.0001274
  4. Khosravikia, F., Mahsuli, M., Ghannad, M. A. (2018). Soil–Structure Interaction in Seismic Design Code: Risk-Based Evaluation. ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering 4(4):04018033. DOI: 10.1061/AJRUA6.0000984. https://doi.org/10.1061/AJRUA6.0000984
  5. Abtahi, S., Mahsuli, M., Ghannad, M. A. (2020). Probabilistic Evaluation of Soil–Structure Interaction Effects on Strength Demands of Shear Buildings. Journal of Structural Engineering 146(1):04019166. DOI: 10.1061/(ASCE)ST.1943-541X.0002458. https://doi.org/10.1061/(ASCE)ST.1943-541X.0002458
  6. Tao, W., Fu, J., Li, Y. (2024). The Possibility of Detrimental Effects on Soil–Structure Interaction in Seismic Design Due to a Shift in System Frequency. Applied Sciences 14(17):7519. DOI: 10.3390/app14177519. https://doi.org/10.3390/app14177519
  7. Zhang, X., Far, H. (2024). Beneficial and detrimental impacts of soil-structure interaction on seismic response of high-rise buildings. Advances in Structural Engineering 27(11):1862–1886. DOI: 10.1177/13694332241255747. https://doi.org/10.1177/13694332241255747
  8. Requena-García-Cruz, M. V., Romero-Sánchez, E., Morales-Esteban, A. (2022). Numerical investigation of the contribution of the soil-structure interaction effects to the seismic performance and the losses of RC buildings. Developments in the Built Environment 12:100096. DOI: 10.1016/j.dibe.2022.100096. https://doi.org/10.1016/j.dibe.2022.100096
  9. Kwag, S., Ju, B., Jung, W. (2018). Beneficial and Detrimental Effects of Soil–Structure Interaction on Probabilistic Seismic Hazard and Risk of Nuclear Power Plant. Advances in Civil Engineering 2018:2698319. DOI: 10.1155/2018/2698319. https://doi.org/10.1155/2018/2698319