Looking at the design all seems relatively well configured. I do notice you have the simulation set to run until you tell it to stop. Without the information regarding how many time steps you ran it is difficult to comment. The reason for this is that with a resonant structure it is critical to get the power out of the resonator and if you do not run the simulation long enough you will not get the expected result.
Do you have a screen shot of the time domain captured by the observation point?
Scott