Help contains everything you want to know...:
Accuracy of the calculated S-Parameter
The S-parameters can only be as accurate as the signals they were calculated from. This means that, theoretically, the S-parameters are only exact if the time signals have entirely come down to zero. Otherwise, information about the system will be lost that would be needed for exact transfer functions.
However, in the simulation, the time signals will never come down. Even when the fields have been propagated entirely through the structure, numerical errors will prevent the signals from reaching zero. At this time, where the signals contain only numerical noise, the simulation should stop because the signals contain no further information about the system. Therefore, a criterion must be found to decide when numerical noise starts to dominate. Also, this criterion should be variable to save simulation time if a lower accuracy is needed.
能把不同的结果放上来大家看看嘛
之前密码想不起来了，今天终于试出来了，附件中是分别设置为30dB和40dB得到的不同结果，期待工程师解答！
精度不同时得到的结果.doc (58 K)
Let see these figures:
-30 dB Accuracy
-40 dB Accuracy
It is quite clear that both of the simulations have the same resonance frequencies at about 4 GHz, 8 GHz, 10 GHz, and 18 GHz and nearly the same S11 level through the whole frequency range. Therefore, the different accuracy level does not affect the simulation results on the electro-magnetic properties.
However, the second figure (-40 dB accuracy) has much "smoother" S11 curve than the first one (-30 dB accuracy) from 4 GHz to 7 GHz. Furthermore, the higher accuracy criterion simulation means lower numerical noise level then improve the S11 resonance performance at 4 GHz and 10 GHz.
According to the above, higher accuracy level need longer field analysis time for the energy to be degraded to the noise level. So, if you check the slover log you should notice that the -40 dB accuracy simulation takes longer duration than -30 dB accuracy simulation.
长见识了，谢解答
域求解微带天线时，为什么solver setting中不同的Accuracy得到的VSWR不同呢？出现这种情况可能是什么原因呢？
如果你把网格再加密的话，采用自适应网格的话，将会得到不同的VSWR，不同的网格设置和精确度肯定得到不同的VSWR，关键是看VSWR的差异，如果变化比较小，那就说明误差已经很小了。
仿真的误差分析在CST的初级教程里面讲的很清楚，建议楼主查找资料看看。
恩，明白了，多谢解释哈
好的，多谢，正在补这些基础知识
声明：网友回复良莠不齐，仅供参考。如需更专业系统地学习CST，可以购买资深专家讲授的CST最新视频培训课程。
