Integral abutment bridges (IABs) are widely used worldwide when the crossing span is short-to-moderate. Due to the elimination of bearings and joints, the seasonal expansions and contractions of IABs are absorbed by the backfill and foundation. However, during the cyclic expansion and contraction of the bridge deck caused by temperature variation, the abutment-backfill-embankment system significantly interacts with the superstructure in a way that is difficult to predict analytically. In the context of the EPSRC-funded PLEXUS project, a 1g small-scale experiment was conducted to assess the distribution of the earth pressures behind the abutment. The experiment comprised a soil box filled with sand. A hinged moveable wall was set against the soil to simulate the integral bridge’s abutment-backfill interaction, and the thermal-induced deformation was introduced by cyclic movements of the wall. This study aims to numerically model the PLEXUS experiments using OpenSees and its visual toolkit STKO, carrying out a numerical-to-experimental comparison as well as a numerical-to-numerical comparison between different finite element (FE) software packages (i.e. OpenSees and PLAXIS, ABAQUS). In the OpenSees model, the soil-wall interaction interface is represented by “EqualDOF” links, and the backfill is considered a discrete domain with “PressureDependMultiYield02” material. The results indicate good agreement between the numerically predicted and the experimentally measured response when the soil is modelled as nonlinear in OpenSees while highlighting specific differences with respect to the different software packages considered. The strategy using “EqualDOF” for modelling abutment-soil interface and plastic “PressureDependMultiYield02” material showed a relatively satisfactory agreement with experimental results.