448 REPORT — 1900. 



Angle of thread. Force required to Area of thread Shearuig force per 



shear thread. sheared. square iucli. 



40'' (1)4,890 lb. -^j:! sq. in. 



(2) 4,760 lb. 

 mean 4,825 lb. 19,080 lb. 



50° (1) 5,400 lb. -2775 „ 



(2) 5,130 lb. 

 mean 5,26.-) lb. is,9r,0 1b. 



C0° (1) 5,500 lb. -300 ., 



(2) 5,600 lb. 

 mean 5,550 lb. 18,500 11.. 



In these experiments only one screw and one nut were used in each 

 pull, the connection to the otlier side of the hexagonal sleeve Ijeing made 

 with a h" steel bolt. Two nuts of each size were sheared. In this case 

 no measurements were taken of twisting torque. 



D. — The screw of 60° was tested on a brass nut -250 inch thick 

 (4 threads) made from hard drawn rod. 



1. Screw broke at 10,280 lb. along line d, iig. 3. 



2. Another similar screw sheared the nut at 10,820 lb. Area sheared 

 ■300 square inch. Shearing force per square inch 36,070 lb. 



In all cases the nut was sheared along the outside surface of the screw, 

 not at the bottom of its own thread, so that the hole left was a tight fit 

 for the screw which had been pulled through. 



Tliese figures suggest the following conclusions : — 



1. That the angle of a flat-ended thread has little effect on the resist- 

 ance of the nut to shearing, except so far as it aflects the area of the 

 surface to be sheared ; and the advantage possessed by the 60° thread 

 over the others is only due to the fact that its flat top is narrower than 

 theirs, and the base of the nut thread correspondingly wider. From 

 analogy with the relative behaviour of sharp and blunted dies used in 

 stamping, it seems that a flat-topped thread with sharp edges should shear 

 a nut more easily than a rounded thread. 



2. That the strength of the thread of the nut, compared with that of 

 the core of the screw, is such tliat generally in practice nuts are stronger 

 than their screws. 



For example, a flat-ended thread of the dimensions of No. B.A. of 

 40-ton steel will break before it strips the thread from a hard drawn brass 

 nut ji^ inch thick. So a similar screw of the dimensions of No. 6 B.A. of 

 the same steel will break sooner than strip a brass nut jV inch thick. 



With steel nuts the nut will generally be very much stronger than the 

 screw. 



3. Considering (a) that the holding strength of a screw bolt is generally 

 determined (and that especially in small screws) by the resistance of the 

 bolt under tensile stress ; and (b) that, as ascertained by Professor Martens, 

 the resistance of a screw bolt to fracture is very largely dimiuished by 

 simultaneous torsional stress ; it is desirable that such resistance as may 

 be desired to tightening or loosening a bolt should be obtained by means 

 of the friction of the under surface of the nut or screw head, and that the 

 friction of the threaded surface of the screw itself should be as small as 

 possible. From this point of view experiments A indicate that an angle 

 of thread of 40° or 50° is to be preferred to an angle of 60°, and that 

 especially so in the case of small screws. 



The authors of this Report are under a great obligation to Professor T. 

 Hudson Beare for his kind assistance in ascertaining the breaking strains 

 of the specimens. 



