About implicit residual smooth
 

I am currently using a 5 stage runge-kutta scheme for my 2-d unstructured 2nd order upwind euler solver. The max CFL number I can use is 2.0, without using implicit residual smooth. When using implicit residual smooth, it can not double the CFL number as many papers said. Especially for naca0012, M=0.8, a=1.25 calculation, the CFL number never exceeds 2.5. My question is: it all these correct?

Thanks for any help. :o