Fluid viscosity in a can and tendency to roll

[oxnume]

How does the viscosity of a fluid in a circular can affect the way the can rolls?
For example, 2 cans of similar size, volumn, and mass are rolled down a ramp with the same slope. The can with a viscous liquid will take more force to begin rolling than a can with a less viscous liquid, but once it gets moving the thicker liquid will roll a longer distance. As a result, on a shallow ramp, the more viscous can will hardly begin to roll and will go a lot shorter than the other one, but on a high sloped ramp, the more viscous can will begin to roll very fast and go a lot further than the less viscous one.
I have tried looking up viscosity but couldn't really find much related.

What are some theories or concepts that can help explain why this occurs?

 

[Chestermiller]

The viscous case can be modeled as a rigid solid. The low viscosity case can be modeled as an inviscid fluid with the same mass, but incapable of causing a torque on the can. So the inviscid fluid does not contribute to the rotational kinetic energy.