[CST MWS] Automotive Antenna Pattern Simulation
I am trying to figure out the most suitable antennas placement on a vehicle for V2X applications, in order to get a good coverage in the area around. The working frequency is 5.9Ghz.
I already have the antenna model designed on CST, together with the farfield radiation pattern. Now, the idea is to place on a 3D CAD structure a set of the antennas (e.g. one for each corner/side) and to obtain and evaluate the consequent radiation pattern.
The problem can be considered as electrically large: the structure is around 200λ. Therefore the Integral Equation seems the more suitable solver available.
Now, how to simulate the total radiation patter of all the antennas working together? My ideas/doubts are the following.
The strategy is to import the original farfield sources and use them as excitation, instead of discrete ports. Is it a good choice? I guess another way could to the placement of the antenna, each one with a discrete port.
Other problem: I found difficult to combine a proper surface mesh setting with the frequency/dimensions I am working with, resulting in very high computational requirements. Is there any chance to get reliable results, simplifying the mesh settings in order to be able to execute the simulation in short times on consumer hardware (say, not within days)?
Is the procedure a correct way of working, or do you have comments? Bad simulator choice? Better to go with the asymptotic solver (ray bouncing)?
It's my first time struggling with these type of problems, so thanks to anyone that is willing to help me.
Welcome prebenl,
You've got the right idea, using the I solver would decrease simulation time because it makes use of surface meshing instead of volume meshing (t solver).
Use the farfield source would be a good start, but I would also consider using the nearfield source, just to see if there's any difference. To use nearfield source, remember to include the geometry where the antenna will be sitting on (in your case, I assume roof of car) and simulate in time domain, capturing the near fields.
It shouldn't take days using I-solver, but you could further decrease the simulation time by making use of symmetry planes in boundary conditions.
Asymptotic solver is another possible option, but depends on what you're after. I don't think A solver is able to produce far-field patterns (I might be wrong). Also, A solver does not support dielectric, but since your problem is a vehicle, I assume they're all PEC for simplification.