This paper considers multiple unmanned aerial vehicles (UAVs) to enable secure and reliable communications between energy-constrained Internet of Things devices (IoDs). First, these IoDs harvest energy using the radio frequency (RF) signals transmitted from a power beacon. Then, the harvested energy is used to exchange information via multiple UAV relays in the presence of an eavesdropper. Among multiple UAVs, the best UAV relay is selected based on the achievable signalto-noise ratio of the first hop. The selected UAV applies a decode-and-forward (DF) operation to forward the information. However, an eavesdropper present in the vicinity of IoDs can wiretap the information transmission through the UAV-to-ground channel. Thus, the UAV employs artificial noise (AN) injection technique to prevent the eavesdropper from intruding on the information transmission. For modeling the composite fading channel, we use the log-normal distribution to characterize the shadowing components. Further, we adopt the Gauss-Hermite quadrature method to derive the closed-form secrecy outage probability (SOP) expression. Moreover, we offer a reliability and security trade-off analysis by providing closed-form expressions for the outage probability (OP) and intercept probability (IP) as reliability and security performance measures, respectively. The numerical and simulation results corroborate the analytical findings, highlight the impact of various channel/system parameters on the secrecy and reliability performance, and provide valuable insights into the system's behavior.Index Terms-Unmanned aerial vehicles (UAVs), wireless power transfer, artificial noise, secrecy and reliability trade-off.