"WAVECIJIDE .\S .\ rOMMrXlC ATIOX MKDUM 



1247 



Table II — Mode C'onversio.n Compahison ok 'JOOO-mc 1.732" 

 Diameter Line with a Hvi'otiietu-al o(),()(X)-mi" 2" Diameter IjIne 



* Difference between major and minor diameters. 



t Separation of guide axes. 



X These are upper-limit values, based on the length of the oval section of pipe 

 which would produce maximum mode conversion, and based on a cross-sectional 

 shape (trifoili which would jjroduce the maximum of mode conversion. 



It is interesting that two of the three ('on\'ersion effects are essentially 

 independent of frequency. Tilt at a waveguide junction introduces a 

 phase-front error and would be expected to cause greater conversion 

 effects at increasing fre(iuencies. We shall see that bends ])roduce a 

 similar mode conversion, also due to a phase front error, that increases 

 with increasing frequency. 



the bend problem 



The problem of transmitting the circular electric wave around bends 

 was recognized as being important at an earlj^ date, and contributions to 

 its solution were made by M. Jouguet,*'^ W. J. Albersheim,^ S. 0. Rice/ 

 and the writer.^ The essence of the problem is as follows: .V bend in a 



Table III — Frequency Variation of Mode Conversions in 2" 



Diameter Guide 



These are upper-limit values, based on the length of the oval section of pipe 

 which would produce maxinuun mode conversion, and based on a cross-sectional 

 shape (trifoilj which would produce the maximum of mode conversion. 



