Nowadays, in material engineering, the use of lightweight components is a basic matter. For developing material properties and also due to the economical and ecological aims, the mass of a structure should be reduced and at the same time its structure quality should be strengthened. Up to now, multiphase steels such as DP (dual phase) steels offer high strength and ductility and have widely used in automotive industry. The increased strength of multiphase steels is the consequence of grain refinement and precipitation hardening by coexisting of softer and harder phases and various grain sizes. The steel used in this study is AISI 5115 that offer impressive mechanical properties such as continuous yielding behavior, high work hardening rate and superior strength–ductility combination which is made it suitable for machine elements. In this study, the dual phase steel was fabricated by a special heat treatment procedure. A series of experiments were then carried out on the produced steel to obtain its mechanical characteristics and to make it ready for photographing in micro dimensions. Next, using scanning electron microscope (SEM), pictures were taken from some regions of the manufactured steel. Two sets of finite element models were then created. The geometry of the first set was created from the SEM images of the material. This method was developed to separate grain boundary from ferrite and martensite phases with a relatively high accuracy. Since the boundaries have significant role in material behavior, it is also need to be recognized between the material phases. At the end of this stage, the pictures were mapped into a matrix involves arrays which show the type of the phases and boundaries. Another method based on DEM is used to create random models. This technique is used a Voronoi type algorithm to construct geometries which are statistically similar to the SEM images. As it is known, ferrite is a ductile phase and fractures with dimpled features due to the void formation and coalescence. On the other hand, martensite is a brittle phase and fractures in cleavage mode. Considering the fracture surface of material, the Gurson model is an appropriate model for simulating the damage inside the ductile material. Therefore, it is a good idea to represent the ferrite and boundary using a GTN model and martensite with elastoplastic behavior. In the next step, parameters of GTN are determined. The representative material volume (RMV) is modeled by two approaches: (1) a unit cell containing a discrete, spherical void at its center, and (2) a unit cell having the same void volume fraction which obeys the Gurson–Tvergaard constitutive relation. The macroscopic stress–strain response and the void growth and coalescence behavior of the voided cell are obtained from detailed finite element analyses and the results show strong dependencies on stress triaxiality and the initial void volume fraction. The micromechanics parameters of the GT model, q1 and q2, are calibrated to minimize the differences between the predicted void growth rate and macroscopic stress–strain relation by the GT model and the corresponding finite element results of the voided RMV. Results of simulations obtained from SEM images shows that the simulation data are in good agreement with experimental results and also analysis of simulation clearly show that the deformation of the material is accompanied by the nucleation of voids that were originally not present in the material. It is showed that voids are initiated from the boundary which is between two crystals commonly with different phases (ferrite and martensite). The obtained results from the finite element model show that the failure always happen in boundaries and then growth over them to form the final failure of the material. Keywords : Dual phase steel, Image processing, Gurson-Tevergaard-Needleman damage model, Microstructure modeling, Failure mechanism