Accuracy of time integration scheme
 

Dear All,

After reading the programmer's guide and prof. Jasak PhD, I'm little bit confused about the formal definition of the order of accuracy of the temporal discretization.

Both reference are confirming that the Crank–Nicolson is second order (programmer's guide page 41 and Jasak page 88).

While for example in the programmer's guide page 37 it's said that the Euler implicit is a first order (which will be used with Crank–Nicolson). Also in Jasak Phd equation 3.38 seems for me as 1st order accurate. This equation is used to get the final discretized equation (equation 3.41) which make the overall order of accuracy inconsistent.

Did I miss something