Ansys HF emag analysis_If you can help

[dadan]

Hello friend, my name is Dado.
I’m a student apsolvent of Microelectronic engineering.
If you can help me , or if you can give me e-mail address of someone who could help me, please do.
I have to do analysis of quality factor and inductivity of the meander line on Si substrate for my seminar work.I’d tried a lot of ways to get good S parameters wich I then use to get Q and L but without success.
My grat concern is:
Boundary conditions and excitation current, how to apply it.In the beggining I'd tried with to me logical steps based on those examples in the «Ansys High Frequency analysis guide», but then I done over fifty diferent analysis based on diferent boundary and excitation conditions, so if you can explain to me wich is the right way, please do.
When I apply current I apply it only to SiO2 dielectric layer, or should I apply it on the Si substrate to.
I apply standard impedance boundary conditions for copper meander line and becouse of that should I eject copper material from the model, all becouse of skin effect wich is significant for cooper but not for Si becouse of their frequency and material properties, but how will Ansys than take into account cpacitive coupling between menaders segments.
Wich distance should bee good enough between the current excitation and port.
I must say that I’d tried diferent combinations related to this including mesh coarsness in all model without success.I'd tried this becouse I did not know what to do after I studied examples in the manuals and I'd tried few basic procedures based on that examples and did not get good results.

I’m working HF electromagnetic analysis in Ansys 10 Multiphysich.
Substrate is 760x760 micrometer, and 400micrometer hight.On top of it there is SiO2 layer of 5 micrometer of thickness, and then goes meander line whose with is 40 micrometer and thisknes is 8 micrometer.
In the begining I model meander line with three elements across skin layer with and ewithout impedance boundary condition on the meander outer surfaces(SHLD= conductivity of copper).
Sio2 Layer I modeled with four elements across thickness, and I thought that I need to apply current excitation on that area under the meander line between the PML and PORT.
Si substrate I modeled with 8 elements across hight witv and without impedance boundaru conditions outer surfaces.Than I allso tried analysis with applied current excitations on the substrate arrea in one or even in two directions allso between PML and PORT.
Hight of the air domen is five times bigger than substrate.
I also applied PEC on the all outer arreas of the model.
For ports I used transmision line ports with a lot of diferent values for reference resistance.
My mesh now approximatively has 20000 elements, but I tried with meshes with three and four time finer meshes and there was no much diference.
I’d also tried to do without copper wich I substitute with impedance boundary condition(SHLD=conductivity of copper0.56e8) on top of SiO2 layer in the meander line shape, and also without substrate that I subsstitute with SHLD for Si=10 and many other combinations as I think I shoul treat this case.
I must say that I use PML layer with very long elements in one dimesion related on the other two dimensions
becouse ANSYS always report me warning message when I use shorter elements than the 1/100 wavelength what would lead to the inacurate results, but I wonder what about aspect ratio for this PML elements wich dont disturb ANSYS becouse it does not send me any warning message about it.
I thought that it would not be so hard to get this parameters but now after all my eforts I start loosing hope so if you can help me and give me right directions, please do.
Here goes one of my command metod examples so that you can see how I treat this case.
I also have a model with PML layer all around my structure like the exmple in ANSYS HF emag guide LINE-FED MICRO PATCH ANTENA, and few more diferent structures wich I used to.In this example PML is very long, but I'd tried with shorter PML and with more elements in it wothout significatn diference.
I hope that I do not ask much from you.
comand cod is in the attachment