The high cycle fatigue life of turbine blades is negatively impacted by high frequency mechanical vibrations caused during operation. One method to mitigate this risk is to use underplatform dampers to dissipate energy from the system and reduce the vibration amplitude. Unfortunately, the state of the art models for such simulations are deterministic, although literature indicates that a large amount of uncertainty exists in measured contact parameters. This uncertainty in the contact parameters leads to significant variations in vibration response. This paper quantifies these uncertainties by considering the input parameters to be stochastic and generating uncertainty bands. A nonlinear solver based on Multi-Harmonic Balance method is used to propagate these uncertainties, and a surrogate model is implemented to increase the computational efficiency. Variance based sensitivity analysis is also performed to rank the importance of each uncertain parameter.