Hi,

For some journals of a numerical analysis nature there are submission guidelines on *how* to report numerical results. If you already know of a few journals where you might want to publish, then check their editorial statements; also, check recent papers in those journals for papers that are of the same nature as the one you have in mind writing. While you are not allowed to copy WHAT other researchers write about, you are allowed to copy/borrow HOW such a presentation is made - at least to some degree.

Hello, 

This paper (and related ones) might provide you an overview about benchmarks and workflows you may go through to grasp ideas on what has been called validation and verification: http://prod.sandia.gov/techlib/access-control.cgi/2007/070853.pdf. Although it is turned to CFD applicants, I think you may have an input. 

Apart this, concerning the tests in numerical analysis, I think the best path you may follow is to look at other papers to find examples that suit to your purposes and journal's guidelines, as mentioned by Michael Patriksson above. Generally, error analyses, mesh convergence tests (grid size, as you told) and reduction of computational effort are some features found in sections talking about numerical results in a paper.

I hope my contribution helps you. ;)

Regards,    

Thanks for the information Patriksson and Gustavo :)