INCREMENTAL SHEATH THH'KXKSS MKASUUEMENTS 303 



The ax'orajic (•ai)acitaiic(' tiom the ptohc clcnicnt to the ^rouiidcd 

 metal core vai'ics Iroin 1.1 to 1 .ii idtuV t'of cahlcs measured. 



3 EXI'EUIMENTAL RESULTS 



S.l (^ircin'l PcrfornKinrc Under StalioiKinj Conditions 



Jnrn nu nt(d (■(ipdcihtncc scnsilirilji for grounded direct capacitance 

 measurements, in normal operating conditions with the probes in contact 

 with a cable sam])le: order of ().()() 1 mmI'- 



Circuil stohiliii/ (uid rrpcdUdn'lili/ for periods over one hour duration: 

 ±0.003 mmF. 



Orcrall Uncdritij of the unbalance indications, as read on the recorder 

 scale within the range of plus or minus 0.25 ^u/uF off center-balance 

 position: ± (3 per cent + 0.003 mmF). 



Mechanical Stahiliftj: Moving or twisting of the connecting leads has 

 no effect on balance stabilit3^ Swinging of the cable under measurement, 

 even beyond the limits encountered in actual working conditions, pro- 

 duces barely noticeable effects on the balance indication. 



Capacitaticc nieasitrement.s on flat poti/elhijlene seimplcs: One of the 

 measuring bridge arms was connected to the auxiliary standard of 1.20 

 mmF. The other arm was terminated by the probe in contact with a flat 

 l)olyethylene sample [placed on a grounded metal-plate. Thickness of 

 samples at the point of contact was measured with a micrometer to the 

 nearest 0.0005 inch. Capacitance unbalance readings were taken directly 

 on the recorder scale to the nearest 0.005 mmF. In order to avoid notice- 

 able "air-gap" and "surface" effects, which occur when stacking several 

 samples, in no case were more than two flat samples in a stack measured. 

 Under these conditions, repeatability of readings was within one recorder 

 division (0.005 ijluF), eciuivalent approximately to one-thousandth of an 

 inch. In a t\'pical case shown on Fig. (i, of 38 measurements taken in the 

 thickness range from 0.052 inch to O.KiS inch, only three measurements 

 were off from the averaging curve by more than 0.002 inch. (Further 

 investigation disclosed that these three points, marked "A" on Fig. G, 

 were all associated with a particular sample.) 



Capacitance tneasurcments on stationary cable samples. In order to 

 estal)lish statistical reliability of measurements on actual cables by the 

 described capacitance method, a numbci- of cable samples were tested, 

 varying in core diameter, a^'(M•age poi.xcthyicne sheathing thickness and 

 mechanical construct ion. 



A typical graph resulting from plotting capacitance increments versus 

 micrometer measun>ments of a cable sample is shown on Fig. 7. Out of 



