Hi, Wolfgang, Thanks a lot. I didn't apply refinement in my code, so there is no hanging node problem. As you said, there is a big difference around the boundary. I set the boundary value always equal to the initial value which is equal to 1. The M*old_solution method gives the correct answer which is 1, but the second method can not. The attached file is the solution.
Best, Xiang On Tuesday, 7 April 2020 16:03:10 UTC-7, Wolfgang Bangerth wrote: > > On 4/7/20 4:58 PM, 孙翔 wrote: > > > > Do you mean the Dirichlet boundary condition? I applied the boundary > condition > > after I form the right-hand side vector. I don't think the boundary > condition > > will lead to different solutions because the right-hand side vector has > been > > different before I add the boundary conditions. > > "I don't think" is the wrong approach :-) You are seeing something you are > not > expecting. In other words, what you "expect" is not happening, and so you > need > to let go of what you think is happening -- you need to *verify* that what > you > think is happening is happening! That's the whole mind set of debugging > problems :-) > > Best > WB > > > -- > ------------------------------------------------------------------------ > Wolfgang Bangerth email: bang...@colostate.edu > <javascript:> > www: http://www.math.colostate.edu/~bangerth/ > > -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en --- You received this message because you are subscribed to the Google Groups "deal.II User Group" group. To unsubscribe from this group and stop receiving emails from it, send an email to dealii+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/89db432f-1bbc-436a-a14c-c79b71d887a1%40googlegroups.com.
