fixing the calculation when polymer not fully mixed
We corrected a bug in the polymer convection term and changed the formulation calculating the injection mobility (assuming polymer fully mixed in wellbore.)
The change basically only affects the results with non-fully mixed polymer. Good agreements are achieved for all the test examples we have and it is not well tested with shear calculation with non-fully mixed polymer (`omega != 1`, we include a suggestion in the PR, while not having an example to verify the suggestion.).
Thanks for the fix, merging!