Derivation of Cyclic Stiffness and Strength Degradation Curves of Sands through Discrete Element Modelling

Fedor Maksimov, Alessandro Tombari

Research output: Contribution to journalArticlepeer-review


Cyclic degradation in fully saturated sands is a liquefaction phenomenon characterized by the progressive variation of the soil strength and stiffness that occurs when the soil is subjected to cyclic loading in undrained conditions. An evaluation of the relationships between the degradation of the soil properties and the number of loading cycles is essential for deriving advanced cyclic constitutive soil models. Generally, the calibration of cyclic damage models can be performed through controlled laboratory tests, such as cyclic triaxial testing. However, the undrained response of soils is dependent on several factors, such as the fabric, sample preparation, initial density, initial stress state, and stress path during loading; hence, a large number of tests would be required. On the other hand, the Discrete Element Method offers an interesting approach to simulating the complex behavior of an assembly of particles, which can be used to perform simulations of geotechnical laboratory testing. In this paper, numerical triaxial analyses of sands with different consistencies, loose and medium-dense states, were performed. First, static triaxial testing was performed to characterize the sand properties and validate the results with the literature data. Then, cyclic undrained triaxial testing was performed to investigate the impact of the number of cycles on the cyclic degradation of the soil stiffness and strength. Laws that can be used in damage soil models were derived.
Original languageEnglish
Pages (from-to)400-416
Number of pages17
Issue number4
Publication statusPublished - 30 Sept 2022


  • DEM
  • liquefaction
  • cyclic degradation
  • damage models
  • one-way input


Dive into the research topics of 'Derivation of Cyclic Stiffness and Strength Degradation Curves of Sands through Discrete Element Modelling'. Together they form a unique fingerprint.

Cite this