next up previous contents
Next: Computing secondary results with Up: Computing Eigenvalues Previous: Computing Eigenvalues   Contents

Adjusting "estimation"

When we now feed gd1 with the edited inputfile (gd1 $<$ doris04.gdf), the end of the resulting output is:
 boundary conditions:
 xboundary= electric, magnetic
 yboundary= electric, magnetic
 zboundary= electric, electric
 
  --------------------
 
    i   freq(i)       acc(i)         cont(i)
    1  503.4599e+6   0.0037654709  0.0027611672       # "grep" for me
    2    1.1576e+9   0.0363792177  0.0338547693       # "grep" for me
    3    1.2000e+9   0.0270761163  0.0465387932       # "grep" for me
    4    1.5669e+9   0.1173770231  0.2662151486       # "grep" for me
    5    1.7357e+9   0.4963611925  1.0000000000       # "grep" for me
    6    1.9216e+9   0.1752955583  0.5479572604       # "grep" for me
    7    2.2211e+9   0.0811233028  0.3181446327       # "grep" for me
    8    2.4446e+9   0.1903593461  0.8610742652       # "grep" for me
    9    2.7054e+9   0.1513682675  0.5798372263       # "grep" for me
   10    3.0470e+9   0.2475596875  1.0000000000       # "grep" for me
   11    3.2898e+9   0.0970269338  0.4556358188       # "grep" for me
   12    3.6312e+9   0.1308660265  0.5174922697       # "grep" for me
   13    4.0755e+9   0.1248055514  0.5934680518       # "grep" for me
   14    4.3644e+9   0.0810408709  0.2603226655       # "grep" for me
   15    5.0381e+9   0.0901497971  0.3161751272       # "grep" for me
 ################################
 # cpu-seconds for eigenvalues :       47
 # start date                  : 30/11/2002
 # end date                    : 30/11/2002
 # start time                  : 13:00:26
 # end time                    : 13:01:32
 # The computation of the eigenvalues has finished normally..
 # Start the postprocessor to look at the results.
 stop .. normal end ..
We see that the results are very bad. The accuracy of all modes is horrible. The reason for this is: We did specify a badly wrong estimation of the highest resonant frequency (We did specify estimation= 10e9). We change the inputfile such that we have
 -eigenvalues
    solutions= 15
    estimation= 2e9   # the estimated highest frequency
    doit
When we compute with the adjusted estimation (gd1 $<$ doris05.gdf), we get as final table:
 boundary conditions:
 xboundary= electric, magnetic
 yboundary= electric, magnetic
 zboundary= electric, electric
 
  --------------------
 
  The first 2 solutions seem to be static and are not saved.
    i   freq(i)       acc(i)         cont(i)
    1  503.4601e+6   0.0000000341  0.0000000180       # "grep" for me
    2    1.0594e+9   0.0000000000  0.0000000000       # "grep" for me
    3    1.1579e+9   0.0000000010  0.0000000060       # "grep" for me
    4    1.2001e+9   0.0000000003  0.0000000038       # "grep" for me
    5    1.2528e+9   0.0000000000  0.0000000000       # "grep" for me
    6    1.5070e+9   0.0000000001  0.0000000002       # "grep" for me
    7    1.5445e+9   0.0000000001  0.0000000017       # "grep" for me
    8    1.5679e+9   0.0000000028  0.0000000893       # "grep" for me
    9    1.5926e+9   0.0000000003  0.0000000026       # "grep" for me
   10    1.6717e+9   0.0000000002  0.0000000021       # "grep" for me
   11    1.7208e+9   0.0000000139  0.0000002458       # "grep" for me
   12    1.7544e+9   0.0000004046  0.0000086559       # "grep" for me
   13    1.7950e+9   0.0000025979  0.0000567143       # "grep" for me
 ################################
 # cpu-seconds for eigenvalues :      102
 # start date                  : 30/11/2002
 # end date                    : 30/11/2002
 # start time                  : 13:04:35
 # end time                    : 13:06:40
 # The computation of the eigenvalues has finished normally..
 # Start the postprocessor to look at the results.
 stop .. normal end ..
These results are accurate enough for now.