stress /rotation curve. In this test, the elongation was not measured 

 because the cable rotation caused rotation of the target of the photo 

 electrical measuring instrument. 



The results of the cable tests are reported in Reference 4. The 

 extensional and rotational response data for the 5/8-inch 1x48 double- 

 armored cable and the jacketed, 1/2-inch, 3x19, torque -balanced cable are 

 also presented in this report in Figures 4 through 11. Also shown in 

 these figures are data predicted by RADAC and RAMSC. The free-end rotation 

 data are an indication of the degree of torque -balancing. Negative 

 rotation is in the direction to unlay the outer wire or strand. The 

 fixed-end test data include the tension-versus -rotation and the torque-versus- 

 rotation curves. The slope of the torque-versus-rotation curve is defined 

 as the rotational stiffness, an important parameter affecting the kinking 

 of a cable. Note that the rotational stiffness for the 1x48 double -armored 

 cable has different values for different directions of rotation (Figure 6) . 

 The cable becomes very soft when the outer layer is unlaying. The rotational 

 stiffness of the 3x19 multistrand cable is nearly constant like a solid 

 rod (Figure 10). 



Program RADAC predicts both the elongation and the rotation of a 

 double -armored cable to reasonable accuracy, but program RAMSC which simulates 

 a multistrand cable can only predict the elongation to the desired accuracy 

 of ±5%. A significant difference was found between the predicted and 

 measured values of end rotation. Specifically, although the magnitudes of 

 the the measured and predicted rotations were in reasonable agreement, the 

 directions of rotation were opposite. The difference was about 14 degree /ft. 

 The predicted rotation indicated a slight overbalance; i.e., the predicted 

 resistance to unlaying was greater than actually occurred. This difference 

 may result from the assumption that the stress distribution of a helical 

 strand is the same as in a straight strand. In reality, wires closer to 

 the center of the cable have higher stresses; therefore, the resultant tension 

 over the strand is closer to the center of the cable, and the positive torque 

 developed by the tension is less. If the average torque developed within 

 each strand is assumed not to change, the total positive cable torque is 

 less than predicted. The predicted rotation will always represent a slight 

 overbalance compared to that of the measured value. The reason manufacturers 

 design 3x19 cables with a built-in overbalance may be to account for such 

 a loss. Other factors that may affect the analytic result include the 

 presence of inter-wire or inter-strand friction, the change in helical 

 radius under load, the jacket stiffness, and the accuracies of input wire 

 dimensions, pitch angles, and elastic properties. 



Program RADAC and program RAMSC have been combined into program TAWAC. 



KINKING CHARACTERISTICS 



Kinking Criterion Formulation 



Both Vachon [6] and Hearle [7] recommend the use of Timoshenko's 

 buckling criterion for a slender rod [8] to predict the onset of cable 



