Meshing paraboloid using UTD
__Hi guy's. Feko suggests very rough mesh for paraboloid (see attached image "Meshed paraboloid"). Paraboloid diameter is very big in terms of lambda (about 500 lambda). A difference between original and meshed surface's confuse me., because distance between them is about several lambda.
__1. Will this affect accuracy? From what surface ray will reflect: from exact paraboloid surface or from big mesh triangle several lambda away from it?
__2. Am i right, that UTD method (which i choosed for paraboloid) also includes GO?
__3. In solver setting - high frequency there is option for UTG/GO "max number of ray interactions".
Am i right, that If, for exhample, max number of ray interactions=3, then FEKO will account for following ray's:
1) reflected- transmitted- reflected,
2) reflected - reflected- diffracted,
3) diffracted - reflected - diffracted,
4) diffracted - diffracted - transmitted,
etc.
__Thank's in advance.
Best Answer
-
UTD and RL-GO in Feko are two different solvers. For reflector antenna applications (which I believe is what you are trying to do), I would recommend to use RL-GO rather than UTD.
1
Altair EmployeeAnswers
-
UTD and RL-GO in Feko are two different solvers. For reflector antenna applications (which I believe is what you are trying to do), I would recommend to use RL-GO rather than UTD.
1 -
Gopinath Gampala_21873 said:
UTD and RL-GO in Feko are two different solvers. For reflector antenna applications (which I believe is what you are trying to do), I would recommend to use RL-GO rather than UTD.
__Thanks for your advice. I just wanted to account for edge diffraction, that's why i decided to use UTD. I thought, that UTD includes GO. It was my misconception.
__I read information that in newer version's of FEKO (2019) it is possblie to account diffraction effect's in GO:
"Edge and wedge diffraction are often dominant in objects such as fins on aircraft. The calculation of these edge and wedge effects following [4] has recently been added to our existing RL-GO high frequency approximation."
__But then a question arises: is there any need in separate UTD solver, if GO solver can do it.
0