________________________________
From: Wolfgang Bangerth <[email protected]>
To: [email protected]
Cc: Pietro Maximoff <[email protected]>; Martin Kronbichler
<[email protected]>
Sent: Monday, August 31, 2009 3:36:19 PM
Subject: Re: [deal.II] What to Use for the Solution Point
> As I refine the mesh (1D), I'd need to scale the right-hand-side
> appropriately as in that case, the solution would grow as 1/h.
If that is the case then you must have forgotten to multiply it with
fe_values.JxW(q)
Check your code and compare it with how step-3, for example, does that.
> Also, how would I go about applying a non-homogeneous Neumann boundary
> condition quickly at the right end? I know that Step-7 shows an example
> but I was just wondering if there was a simpler or another way.
That's the recommended way to do it.
Best
W.
Cheers Wolfgang. That was precisely the problem. I'm so relieved. I forgot to
multiply a part of my stiffness matrix by fe_values.JxW(q_point). Now, with
quadratic elements and 4 mesh refinements, the solution is practically
identical to the analytical one.
Now, I'm off to study example 7.
Best regards
Pietro
_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
