“…The simulation models used in previous studies differ from each other in many ways and can be classified into distinct groups based on the model features: (1) the type of analysis: linear or nonlinear-that can significantly affect the response of the structure as documented in Marjanishvili and Agnew (2006), where a comparison between different analyses shows that the predicted responses can vary significantly when performing static/dynamic and linear/nonlinear analyses, (2) the typology of elements: continuum, beam-column elements or a combination of both have been successfully used for modeling local and global phenomena of progressive collapse: examples of micro-models can be found in Khandelwal and El-Tawil (2007), Sasani and Kropelnicki (2008), Kwasniewski (2010), and Bao et al (2008), while examples of macro-models are those utilized in Kaewkulchai and Williamson (2004), Bao et al (2008), and Bao and Kunnath (2010) and an example of the use of hybrid models is reported in the work of Alashker et al (2011), (3) the dimension of the model: planar or three dimensional-that is crucial in capturing spatial effects: most of the published work were conducted on two-dimensional structures (Khandelwal and El-Tawil 2007;Bao et al 2008;Kim et al 2009) and very few on three-dimensional models (Ruth et al 2006;Alashker et al 2011) making the comparison of results very challenging, (4) the floor system, modeled using a collection of beam-column, shell or brick elements , which plays a key role in determining the response of three-dimensional structures, as shown by Yu et al (2010), Alashker et al (2011) and Li and El-Tawil (2014), (5) the loads applied on the structure, which can include only the self-weight or the self-weight and a fraction of the design dead/live loads.…”