Fixing an monopole simulation in CST
Hello everybody!
I'm trying to fix the results on my simulation, cause they're too far from what I was expecting.
I'm simulating an monopole in CST. It have an length of 220nm and 20nm of radius. The frequency range is from 100THz to 600THz.
I'm using the transient simulation in -80dB, and using an probe to calculate the E-field in the x direction (picture 1).
What is happening is that the signal that the probe measured have to many "drizzle". It have to show me a picture like the picture 2.
My professor said that it may be the mesh configuration. So if anyone have an idea pleas help me.
Thank a lot.
you could check on the mesh definition view. Maybe you could adjust the lines per wavelength to the high. and the equil~ ratio to the high. i forgot the name. sorry.
I think, if you get the mesh high than 1 million, it would be more accurate. And also, if you use the transient solver, you should adjust the accuracy to lower. thanks.
When you talk about the mesh higher then 1 million, you're mean the mesh line ration limit? And about the equil~ ratio I don't know what you're saying, cn you be more specific. And about the llines per wavelenght I put 110, the same to the lower mesh limit, cause if I get a higher number it says that it's out of memory.
Other adjustes that I made is in the maximun number of pulses that I put 100, and in the refine at PEC/lossy metal edges along coordinate axes only that I put 5. So I want to know if this adjustes will help my simulation.
Thanks for the reply.
sorry for the latest reply, i've doing a lot of simulation here. And i haven't open the edaboard until these weeks
It means:
1. lines per length --> make it higher
2. Click the mesh properties. and chose the specials, and find the "refine at PEC/lossy metal" to make it higher than 5.
I am sure that will help you...
So you are not happy about the ripples in the received signal?
What is the distance of the probe to the structure? This is important about setting the maximum number of pulses (i.e. simulation duration). You may not need 100 pulses if your probe location is close to your antenna. You can also use symmetry planes (in your case if your monopole is just a simple monopole, you may be able use two H symmetry planes, which will reduce the computation domain (i.e. the number of unknowns) by 4).
Also check your energy. Does it start to have ripples towards the end? Your energy first should go up (during the excitation) and then should monotonically go down (it may have different decay slopes, but it shouldn't increase at all after the excitation)
My suggestion would be to change PML settings. Default is 0.0001 reflection with 4 layers I guess. Change it to about 10 layers, with 1e-8 or 1-9 reflection.
