Yes. You can define it as a force if you want longitudinal stress. Or a pressure on the inside if you want hoop stress. (P_applied = S*ID / (2*T), T = wall thickness, S = stress; check me).
I'll bet you it doesn't change the frequency significantly, compared to no stress. Let me know if I'm wrong.
You are correct Mike. Only tensile or compressive loads will or could change the frequency results significantly. This was taken from the help file...
Effect of Loads on Frequency Analysis
When building the geometry of a model, you usually create it based on the original (undeformed) shape of the model. Some loads, like the structure’s own weight, are always present and can cause considerable effects on the shape of the structure and its modal properties. In many cases, this effect can be ignored because the induced deflections are small.
Loads affect the modal characteristics of a body. In general, compressive loads decrease resonant frequencies and tensile loads increase them. This fact is easily demonstrated by changing the tension on a violin string. The higher the tension, the higher the frequency (tone).
You do not need to define any loads for a frequency study but if you do their effect will be considered.
To include the effect of loading on the resonant frequencies, you must use the Direct Sparse solver. If your Solver option is set to Automatic, the Direct Sparse solver will be used if loads are defined for a frequency study.
Reading that, it sounds like I'm wrong.