On 10/1/25 05:39, HC Zhang wrote:
*
======================= RT1-DGQ1 ========================
cells uL2 uH1
pL2
4 2.067e-01 - 3.128e+00 - 2.055e+00 -
16 5.324e-02 1.96 1.596e+00 0.97 4.447e-01 2.21
64 1. 307e-02 2.03 7.968e-01 1.00 1.170e-01 1.93
256 3.256e-03 2.01 3.983e-01 1.00 4.811e-02 1.28
1024 8.195e-04 1.99 1.997e-01 1.00 2.965e-02 0.70
4096 2.086e-04 1.97 1.005e-01 0.99 2.049e-02 0.53
16384 5.409e-05 1.95 5.090e-02 0.98 1.444e-02 0.51
======================= RT2-DGQ2 ==========================
cells uL2 uH1 pL2
4 3.405e-02 - 6.559e-01 - 4.237e-01 -
16 4.760e-03 2.84 1.666e-01 1.98 6.974e-02 2.60
64 7.869e-04 2.60 5.200e-02 1.68 4.539e-02 0.62
256 2.285e-04 1.78 2.778e-02 0.90 4.671e-02 -0.04
1024 8.213e-05 1.48 1.923e-02 0.53 3.686e-02 0.34
4096 2.966e-05 1.47 1.371e-02 0.49 2.707e-02 0.45
16384 1.061e-05 1.48 9.751e-03 0.49 1.944e-02 0.48
If helpful I can post minimal code snippets (the weak form assembly, the
boundary projection call, and the lines where I subtract mean_pressure)
and the convergence tables/logs. Thank you again for any guidance.
So if I see this right, the solution actually *does* converge, just not
at the expected rates. That's a different situation than "solution does
not converge" because it means that the function that projects the
boundary values must be doing "something" right at least -- it's not
clear whether it's doing everything right, but it's also at least not
completely wrong :-)
So start to poke some more. Can you pick boundary values that are in the
discrete space? In that case, the discrete boundary values are exact (or
should be) and if you then get the correct convergence rates, you know
that the difference in behavior is due to having exact vs projected
boundary values. If you still get wrong convergence rates, you can
suspect that the problem lies elsewhere.
Best
W.
--
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 [email protected].
To view this discussion visit
https://groups.google.com/d/msgid/dealii/a2721e50-461f-437e-b981-834416f80014%40colostate.edu.
