Yes, I was using the direct solve (-pc_type lu) in both cases. On Wed, Dec 30, 2015 at 4:31 PM, Paul T. Bauman <ptbau...@gmail.com> wrote:
> > > On Wed, Dec 30, 2015 at 5:27 PM, Vikram Garg <vikram.v.g...@gmail.com> > wrote: > >> I made a test case for debugging a potential bug in the Adjoint Refinement >> Error Estimator. However, I am encountering a somewhat bizarre issue on >> running it. >> >> |F| = 3662.286105743243 >> |J| = 1137.777777779091 >> * 0 KSP Residual norm -nan * >> > > That's your problem. You're getting a NAN so the solver is not doing > anything. What solver options are you using? I presume this is with PETSc? > What PETSc options? > > Note that with PETSc >= 3.6.x, some change in behavior has been noted that > is still being tracked down. Did you try direct solvers? > -- Vikram Garg Postdoctoral Associate Predictive Engineering and Computational Science (PECOS) The University of Texas at Austin http://web.mit.edu/vikramvg/www/ http://vikramvgarg.wordpress.com/ http://www.runforindia.org/runners/vikramg ------------------------------------------------------------------------------ _______________________________________________ Libmesh-users mailing list Libmesh-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libmesh-users
