Monday, September 15, 2014

P2 moment fitting

The problems appear to be with boundary triangles that are very small compared to mesh size.




Old method with $h^{-1}$ subtriangles fitting the boundary

h        Nodes   Time      |   L2(Gamma)            H1(Gamma)    
---------------------------------------------------------------------------------------
0.1000      10    0:0:1s    |  1.44e-03                8.17e-02
0.0500      20    0:0:4s    |  1.93e-04   2.91      2.25e-02   1.86
0.0250      40   0:0:16s   |  2.56e-05   2.91      5.84e-03   1.95
0.0125      80    0:1:2s    |  3.28e-06   2.96      1.49e-03   1.97
0.0063     160   0:4:25s  |  4.14e-07   2.99      3.73e-04   2.00
0.0031     320  0:22:13s |  5.21e-08   2.99      9.38e-05   1.99
----------------------------------------------------------------------------------

Moments fitting P2
\begin{array}[h] {lrr|rr|rr} h & d.o.f & time & L_2(\Gamma) & & H_1(\Gamma) & \\ \hline 0.1000 &    10 &    0:0:1s &  2.58e-03  & & 8.16e-02 & \\0.05000 &    20 &    0:0:2s &  7.33e-04 & 1.82 & 2.29e-02 & 1.84 \\0.02500 &    40 &    0:0:4s &  1.47e-04 & 2.32 & 9.77e-03 & 1.23 \\0.01250 &    80 &    0:0:8s &  5.25e-05 & 1.49 & 7.82e-03 & 0.32 \\0.00625 &   160 &   0:0:15s &  8.28e-06 & 2.66 & 1.66e-03 & 2.24 \\0.00313 &   320 &   0:0:31s &  2.47e-06 & 1.74 & 1.05e-03 & 0.67 \\0.00156 &   640 &    0:1:2s &  1.82e-06 & 0.44 & 1.75e-03 & -0.74 \\0.00078 &  1280 &    0:2:6s &  2.34e-07 & 2.96 & 4.21e-04 & 2.06 \\0.00039 &  2560 &   0:4:14s &  2.81e-07 & -0.27 & 9.64e-04 & -1.20 \\0.00020 &  5120 &   0:8:42s &  2.47e-06 & -3.14 & 2.11e-02 & -4.45 \\0.00010 & 10240 &  0:19:22s &  1.82e-06 & 0.45 & 4.93e-02 & -1.22 \\\end{array}


No comments:

Post a Comment