[官方]cst官方网站关于频域求解的常见问题（英）

[sell=1]When is it advisable to use the Frequency Domain solver? (Faq #35) ofVEao  
(Last modified: 9 Feb. 2006)  Add to My FAQs  P4i3y{
$V  
  w<v1N  
n#)PvV~  
In particular for lower frequencies, i.e., the total dimensions of the simulated structure are smaller than approximately one hundredth of the wavelength, the frequency solver can have advantages. A second indication to use a frequency domain solver is the presence of a Floquet boundary with a certain phase shift between opposite boundaries. C+T&O  
R&]c"cO L8  
Especially the combination Frequency Domain approach and tetrahedral meshing is often used for cross-checking results produced by the Time Domain solver on hexahedral meshing (and vica versa), since both simulation techniques are completely different. So the idea of CST's complete technology approach is to use the solver/mesh that fits best for the majority of the simulations and to use the less efficient one for the verification of the results. Q +R3H
,  
u0wn=Dg  
;gdi=>S_  
Ck:#1-t8{  
top  ?VU
gwP_=  
>7"$}5d  
Direct or iterative solver, which one to choose? (Faq #109) r(pwOOx  
(Last modified: 22 Sep. 2004)  Add to My FAQs  =rtS#u Y  
  QCVsVG!sN  
poTl|y @  
In general, the direct solver is recommended for relatively small problems with multiple excitations. The iterative solver is advantageous for large problems. +}kO;\  
]Jja  
U.HoFf+HN  
OiYNH~hv  
top  \nrgAC-b  
Fecx';_1`  
Specialized solvers for resonant structures (Faq #116) \FIa,5k8  
(Last modified: 9 Feb. 2006)  Add to My FAQs  w9/nVu  
  R|nEd/'<  
r9+E'\  
The new Resonant solvers, available within the Frequency Domain Solver Dialog, are extremely powerful if the structure under investigation has a very high Q. In case you are only interested in the S-parameter information, the "Resonant: Fast S-Parameter" is recommended. If also the 3D field information is required (eg. for power loss calculation), the solver "Resonant: S-Parameter, Fields" should be chosen. A specialty of this solver is, that field patterns can be obtained even in the postprocessing, without having defined a monitor before. This option is available within the S-Parameter 1D plot, where an axismarker can be placed and via Right mouse (context menu) the 3D E- and H-field at this frequency can be calculated. (s5<  
Y]hV-_2+Do  
dK[*
 
ROPC |  
top  @][ a8:Y9I  
pk;ffq@  
How can I set multiple adaption frequencies for the TET solver? (Faq #120) I93 ~8wQ  
(Last modified: 9 Feb. 2006)  Add to My FAQs  =X)Q7u".7  
  20$F$YYuk  
)!'Fa_$ e  
It is possible to manually set multiple adaption frequencies (e.g. for a 2-arm diplexer) via VBA (copy lines in the history list). In the following example the mesh adaption will run at 1.3 AND 2.2 GHz, if GHz is specified as frequency unit, creating a good grid for the whole frequency range. -J-3_9I  
F_ ,L2J  
With Solver pp*bqY  
.FDSolverAddFrequencyIntervall "2.2", "", "1", "Adaptation" Rrh<mo(yj#  
.FDSolverAddFrequencyIntervall "1.3", "", "1", "Adaptation" /#:Rd^  
end with %~][?Y ><  
5{8x*PSl  
Access to this feature will be more comfortable in future versions. JZWgr&O<  
^fFtI?.6jI  
~gSwxGT7d  
cWgbd^J  
top  i_[^s:*T  
V"w`!  
What are "Bad elements" in the Frequency Domain solver, and how do they influence the solution? (Faq #223) *?EO n-  
(Last modified: 11 Jan. 2006)  Add to My FAQs  i,mrMi c#  
  sI^@A=.@  
-3C* P  
The so-called "Bad elements" may occur during the tetrahedral mesh generation in the CST MWS frequency domain solver. Very "flat" or very "thin and high" tetrahedra are typical examples. R;0W+!fE  
x>U1t!'  
The presence of bad elements can influence the solution in two ways: ox!|)^`$_  
4 *Bp  
- Influence on the mesh adaptation process: Bad elements in the initial mesh can negatively affect the mesh adaption process. During the mesh refinement in the mesh adaptation, more refinement steps are typically needed in order to obtain a good final solution. X_vI0YX9  
I;`Ko_i  
- Influence on the solution's accuracy: Bad elements which also have large dihedral angles worsen the condition number of the matrix. This typically leads to larger numbers of necessary iterations in the iterative solver, and may also lead to less accurate solution. YRg=yVo2  
,[Dh2fPM,  
What to do: the user can try to improve the quality of the initial mesh by adjusting the mesh parameters: S4#A#a2J  
R59
e&   
1) Local mesh parameters (select object, then right mouse button -> Mesh properties): in the areas in which bad elements occur, you can choose a smaller mesh step for the individual objects concerned: this typically leads to better mesh quality. 1mT|o_K{ T  
} l:mN  
2) Global mesh parameters (Mesh -> Mesh properties): if you choose a larger number of steps per wavelength, as well as "Smooth density transitions" in t .. :5TXA  
f28gE7Y\a  

未注册仅能浏览部分内容，查看全部内容及附件请先 登录 或 注册

?????????????? }V1DyLg:  
堪布了啊

每问题呀。。。。。。。。。。。。。。。。。。。。。。。。。。

请问CST官方网址在哪里呀~~~谢谢啦

www.cst.cn

谢谢楼主啦，看看先

很久前的了吗？

cst官方网站关于频域求解的常见问题（英）

恶趣味完全厄齐尔全文

谢谢楼主啦，