SOME PHYSICAL PROPERTIES OF WIPING SOLDERS 107 



to this paper, and shows that all runs could be fitted without significant 

 variation by an equation of the form of Equation 1, and that the values 

 of b for these different runs did not differ significantly from one 

 another. On the other hand, the values of k thus obtained did differ 

 significantly from an expression of the form k = K{W°'/V''), in 

 which K, a, and c were selected to give the best agreement w^ith the 

 data, indicating that Equation 2, as applied to this material, is not 

 strictly correct. The values of a and c thus obtained differed con- 

 siderably from the values of n and (3« + l)/2 respectively that would 

 be required if Equation 4 applied, n being given by equating b to 

 5{n + l)/2 in accordance with this equation. On the other hand, 

 these differences were not indicated as necessarily significant, so that 

 it appears that Equation 4 applies to these data to a fair approxima- 

 tion, the divergence being perhaps due wholly to the approximations 

 employed in the theoretical development. 



Regardless of the accuracy of Equation 4, the considerations dis- 

 cussed above indicate quite clearly that the flow of solders in the wiping 

 range is quasi-viscous (independent of strain, and of a yield point 

 requirement) and hence similar to the flow of many colloidal solutions. 

 This conclusion is in agreement with the fact that at these tem- 

 peratures solder consists of a solid phase dispersed in a liquid phase. 

 It is furthermore apparent that the quantity b- 5 (or of 1 - n, if use is 

 made of the theory given) is a measure of the departure of the material 

 from a strictly viscous condition. 



Evaluation of Floic Constants 



With the exception noted above, the runs made with the various 

 solders tested were confined to a single sample size and a single load. 

 In each case values of b were obtained by plotting log dh/dt vs. log h, 

 values of the former quantity being found by plotting tangents as 

 described. To illustrate the reproducibility of points in the plot of 

 log rate vs. log height. Fig. 2 has been prepared from Fig. 1. Curves 

 I and II of Fig. 1 were plotted to different scales of sample height 

 and the tangents drawn. In Fig. 2 are plotted values of log dJi/dt 

 against log h from the two curves. It is seen that the line would be 

 drawn in practically the same place if either set of points was taken 

 alone. 



In comparing solders it is evident that the sample height after a 

 given time is a rough measure of relative consistency, the softer 

 material showing the lower sample height. But from Equation 1 it is 

 evident that such a comparison may be misleading for materials 

 showing different values of b. This is brought out in Fig. 3, in which 



