Detecting internal defects such as fatigue cracks in their early stages is critically important to avoid catastrophic failures. However, detection of micro-scale defects poses a challenge to NDT/SHM (Non-Destructive Testing & Structural Health Monitoring) community. Conventional linear ultrasonic techniques that use absolute time of flight and attenuation to monitor damage progression encounter difficulty in detecting small defects. On the other hand, nonlinear ultrasonic techniques have been proven to be more reliable and sensitive to micro-scale defects. This study uses the nonlinear Sideband Peak Count Index (SPC-I) technique to evaluate the nonlinearity of the material. The problem is simulated using the finite element method (FEM). Fatigue cracks are simulated as breathing cracks. The problem is modeled as an infinite medium to minimize the geometrical effects such as reflections from the boundaries. A narrow band signal is excited and propagated through the material in a single sided excitation/detection setup. The degree of nonlinearity caused by breathing cracks is investigated for multiple configurations. First, the problem is modeled with no cracks to be taken as a reference condition. Then thick crack and thin crack (breathing crack) are modeled, analyzed, and compared. Finally, problems with different number of breathing cracks are simulated and analyzed. All simulated results are compared to investigate the dependence of the degree of nonlinearity on the density and orientation of the cracks.