hfss scan angle array
Hi,
U and V vectors define the 'lattice' geometry of the array. For example, for the case at hand, I guess you want to design a uniform 1D array of patches. So, you will have N elements along the y direction (N columns) and exactly 1 row.
Classical array rectangular lattices require U and V vector to be orthogonal with each other. This means that the correct settings for 'Directions' could be U-->X=1, Y=Z=0 and V-->X=Z=0, Y=1. You can use the same settings for 2D rectangular arrays, while for alternative configurations, such as skewed trialgular lattices, U and V mustn't be orthogonal (e.g. define U and V the way they form a 45 deg angle, U-->X=Y=1, Z=0, V-->X=Z=0, Y=1).
Distance between cells should be the same, in the case of rectangular lattice, of the µstrip contained into the airbox. You have to choose this disctance keeping into account for mutual coupling due to surface currents (if the cell is too small) and grating lobe formation (if the cell is too large). I suggest to choose a cell having a 0.75λ0 _ 0.85λ0 period (distance between cells) along V/y direction for broadside arrays. In the case of 1D array at hand , the cell size along U should be larger than V, in order to reduce losses due to dielectric fringing effects. In a few words, try with a rectangular unit cell. Otherwise, for a 2D regular array, you will choose a square cell, if radiation symmetry is required, since ther's have a periodicity along both X and Y.
For the uniform linear array at hand, number of cells must be 1 along U, and N along V.
About the phase shift, you should know from array factor theory that the direction of radiated beam depends on the relative phase of feeding currents between two adjacent elements. If you want a uniform phase broadside array, the differential or relative phase shift must be 0 along the direction/s of array, so as a constructive interference of the powers radiated by each element occurs in the direction normal to the surface. Alternatively, instead of definying phase shift directly, you can define array shift indirectly by setting the desired scan angle of the array.
I believe this tool is a good first way of analysis of the array but, in a following stage, for finite arrays, the simulation of both the array and the feeding network must be performed.
Moreover, HFSS allows the simulation of infinite arrays using Master-Slave boundaries and Floquet's ports.
Design of arrays, anyway, involves a lot of problems and aspects, it would be advisable to have a complete knowledge of array theory.
Hope this helps.
Added after 10 minutes:
By the way,
instead of plotting the array factor, which is essentially the radiation pattern of an array of isotropic sources, you can just plot the directivity along the direction of the input current (x, so X_Directivity) of the array of patches for φ=90deg.
Regards.