Purpose -The purpose of this paper is to establish a model-based framework allowing the simulation, analysis and optimization of friction stir welding (FSW) processes of metallic structures using industrial robots, with a particular emphasis on the assembly of aircraft components made of aerospace aluminum alloys. Design/methodology/approach -After a first part of the work dedicated to the kinetostatic and dynamical identification of the robotic mechanical system, a complete analytical model of the robotized process is developed, incorporating a dynamic model of the industrial robot, a multi-axes macroscopic visco-elastic model of the FSW process and a force/position control unit of the system. These different modules are subsequently implemented in a high-fidelity multi-rate dynamical simulation. Findings -The developed simulation infrastructure allowed the research team to analyze and understand the dynamic interaction between the industrial robot, the control architecture and the manufacturing process involving heavy load cases in different process configurations. Several critical process-induced perturbations such as tool oscillations and lateral/rotational deviations are observed, analyzed, and quantified during the simulated operations. Practical implications -The presented simulation platform will constitute one of the key technology enablers in the major research initiative carried out by NRC Aerospace in their endeavor to develop a robust robotic FSW platform, allowing both the development of optimal workcell layouts/process parameters and the validation of advanced real-time control laws for robust handling of critical process-induced perturbations. These deliverables will be incorporated in the resulting robotic FSW technology packaged for deployment in production environments. Originality/value -The paper establishes the first model-based framework allowing the high-fidelity simulation, analysis and optimization of FSW processes using serial industrial robots.