Driven by the need of robust, cost-effective, and ready-to-use solutions to connect wirelessly thousands to million of nodes, an increasing number of applications (e.g., Smart Grid, IoT, Big Data) use large-scale Wireless Mesh Networks as transmission support. Tools and methodologies to study the performance of such systems are constantly sought and, in particular, they become fundamental in the feasibility assessment of the high number of possible applications. In this paper, a simulation tool is proposed to study the performance of a particular kind of wireless mesh networks, based on the RF-mesh technology. The tool was used in the context of a large-scale smart grid AMI (Advanced Metering Infrastructure) system. The modular nature of the implemented tool allows a smooth extension to the performance analysis of other types of Wireless Mesh Networks using technologies similar to RF-mesh. The tool, coded in Java and Python, considers different types of traffic and provides the packet collision probability, the end-to-end delay, and several other performance indexes of large scale (i.e., up to $6$ thousand nodes) instances in a reasonable time.