The numerical analysis of the reinforced concrete beams strengthened with CFRP is presented. The beams
previously tested experimentally under long-term loading are selected for numerical simulation. The numerical modelling is
performed by evaluating the beam’s work at various stages: the work stage before the long-term loading period, the work
stage under the long-term load action, the work stage when the external load is removed and the work stage until failure. The
work stages of all modelled beams are described in more detail. To analyse the behaviour of beams at different work stages,
the numerical modelling using the phase analysis is performed. Different finite element groups are evaluated in each phase of
analysis. The external load is increased, maintained and reduced. The finite elements of the CFRP layer are activated at a
certain work stage for evaluating the strengthening effect. To assess the accuracy of the numerical analysis, each beam is
modelled from the finite elements of various sizes. The paper presents the process of the numerical modelling and the
predicted deflections. The numerically predicted deflections are compared with the deflections of the experimental study. The
modelling of the behaviour of the strengthened beams has shown that the nature of the long-term deflection differs from that
obtained in the experiment. The increment of the numerically predicted deflection decreases gradually over the long-term
period. Meanwhile, the experimental long-term deflection increment is characterised by the sharp increase and decrease at
the start of the long-term period. This contradiction shows that the experimental long-term deflections are greater. However,
over time, the numerical model deflections may reach and exceed the experimental deflections due to steady increase. The
smaller size of the finite elements causes the increase in the cracking moment and the higher moment when the yielding of
the tensioned reinforcement occurs. However, the cracking moment obtained by the numerical modelling is much higher
than that obtained by the experimental modelling. However, when the yielding strength of the tensile reinforcement is
reached, the considered moment is smaller than the experimental one.