How can two linearly polarized Yagi antennas be used to produce circular polarization?
Why Circular polarization is just two linear waves at right angles to each other whose fields are 90 degrees out of time phase, so the resultant field vector rotates one full turn each RF cycle. To build that with Yagis you mount two identical arrays on the same boom axis with one set of elements horizontal and the other vertical, driven elements at the same boom position, and feed them in quadrature (commonly by adding an extra quarter wavelength of line to one). Which one leads by 90 degrees determines whether the result is right-hand or left-hand circular, and this crossed-Yagi arrangement is standard for satellite work because it tolerates spacecraft spin and Faraday rotation.
Watch out Stacking two Yagis in parallel planes keeps both sets of elements in the same polarization, so phasing them only steers or shapes the pattern (in phase gives stacking gain, 90 degrees gives a beam tilt); you never get rotation without spatial perpendicularity.
Circular = crossed plus quarter: elements 90 degrees apart in space, feed 90 degrees apart in phase.