for an unusual mix or an unusual alignment of factors necessary for the occurrence of the phenomena. From the viewpoint of optimizing efficiency, the computer simulation program will become very valuable in the future, because it is possible to change one parameter and observe its effects upon efficiency while the other parameters remain constant. This is often difficult to do experimentally if we are dealing with a resolution of 0. 1% in efficiency. It is also probable that the chance alignment of parametric factors becomes more important as efficiency refinements deal with small improvement in efficiency, say from 91% to 92%. Obviously such computer simulation becomes very valuable in a quality assurance program where the objective is to maintain efficiency within very tight tol- erences and above all to avoid highly destructive mechanisms. In 1975 Dr. Joseph Nahas ' ’ 7 reported upon Math modeling and computer simulation that he had successfully carried out for a rectenna element modeled after the standard RXCV rectenna element in many respects. The contract for which this final report is being written was issued in 1975 and contained a task in which the RXCV element was to be math modeled and its operation simulated with a computer program. This aspect of the contract was assigned to E.E. Eves of Raytheon. He was assisted by Phil Knight. They requested and obtained a copy of Nahas' s computer simulation program, which was of considerable aid to them. Although Nahas' s basic approach to the problem was very similar to that which had been taken by Eves in the earlier time period ' Dr. Nahas' s program was not used directly or modified for two reasons. It did not directly model the RXCV element, and the detail of the math modeling was so great that the computer time to obtain performance data represented an expense that the present contract could not support. Further, we wanted to compare the results of the computer simulation with experimentally measured values of diode dissipation and the latter could be obtained only with the use of the ground-plane fixture which omits a consideration of the half wave dipole input. Hence, the modeling of the antenna was eliminated. This provided considerable simplification. The skin losses were modeled only for the fundamental frequency and the number of filter sections to be modeled was two rather than five in the Nahas model. This represented additional simplification. However, every effort was made to simulate the diode characteristics and function very accurately, starting with the most basic characteristic of the junction itself. Further precautions were taken in setting up the program to eliminate computer errors at points of sharp discontinuity. Predictor-corrector routines were incorporated as they were in the Nahas program. Both the Nahas and the Eves-Knight computer simulation programs depend upon watching the buildup of the current and voltage waveforms from a transient to a steady state condition in the rectenna element after the step application of microwave power input. The convergence to a steady state condition for the Eves-Knight program requires about 20 cycles. The number of steps within a cycle is 8192 and a printout is made of designated parameters 128 times at equal intervals in the last cycle. The computer time required to obtain the results from one set of input parameters is 3 minutes on a CDC6700 computer. The cost is about $ 20.00 per set of data.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==