Coherent optical frequency domain reflectometry (c-OFDR) enables continuous and automated crack monitoring in concrete structures due to its high spatial resolution and long sensing range. However, the selection of a suitable DFOS/adhesive combination (DAC) remains a key challenge and is typically based on empirical knowledge. As a result, the monitoring quality is highly dependent on the user's experience. Inappropriate choices can result in undetected cracks, degraded signal quality, miscalculated crack widths, or even fiber breakage. This study proposes a semi-analytical model for predicting strain curves for arbitrary crack patterns, based on experimentally derived parameters, such as sensitivity, strain lag parameter, and the extent of bond disturbance due to transverse cracking. The model was validated using strain measurements from tests on reinforced concrete specimens with multiple cracks. With only a few input parameters, it enables accurate representation of crack-induced strain peaks and allows for the consideration of measurement limitations typical for c-OFDR systems. Based on this, design charts and an open source software tool were developed to support practitioners in the systematic selection of DACs, tailored to the expected crack pattern.