In this paper, we investigate the dynamics of a nonlinear delay differential equation model for passive mode-locking in semiconductor lasers, when the delay model is seeded with parameters extracted from the gain and loss spectra of a quantum dot laser. The approach used relies on narrowing the parameter space of the model by constraining the values of most of the model parameters to values extracted from gain and loss measurements at threshold. The impact of the free parameters, namely, the linewidth enhancement factors that are not available from the gain and loss measurements, on the device output is then analyzed using the results of direct integration of the delay model. In addition to predicting experimentally observed trends such as pulse trimming with applied absorber bias, the simulation results offer insight into the range of values of the linewidth enhancement factors in the gain and absorber sections permissible for stable mode-locking near threshold. Further, the simulations show that this range of permissible values progressively decreases with increasing bias voltage on the absorber section. This is important for telecomm and datacom applications where such devices are sought as pulsed sources, as well as in military RF photonic applications, where mode-locked diode lasers are used as low noise clocks for sampling.