[Moon11]

I have made a first-order F.V. (upwind scheme) code for the shallow water system in 1 D. I have found that when I increase the value of CFL until 0.99, the resolution of the solution gets better special in the dam break problem and the method seems to be more than first-order method!!, my question is can cfL increase the accuracy of the method? or this is due to the solution evolving more rapidly, causing it to be artificially steep in the numerical solution.

[FMDenaro]

Originally Posted by Moon11

I have made a first-order F.V. (upwind scheme) code for the shallow water system in 1 D. I have found that when I increase the value of CFL until 0.99, the resolution of the solution gets better special in the dam break problem and the method seems to be more than first-order method!!, my question is can cfL increase the accuracy of the method?

Not the accuracy order, but it can locally reduce the amount of numerical viscosity

[Moon11]

Originally Posted by FMDenaro

Not the accuracy order, but it can locally reduce the amount of numerical viscosity

Thank you for your quick response... Does this mean that it would be better to choose a high value for CFL as long as the chart is stable?

[FMDenaro]

Originally Posted by Moon11

Thank you for your quick response... Does this mean that it would be better to choose a high value for CFL as long as the chart is stable?

Yes but the CFL is local, it depends on the velocity field and step size. Thus, you cannot suppose to have a uniform high value everywhere.