44G 



REPORT — 1900. 



bi"ass nut, at a strain much below the calculated breaking strain. The 

 form of the fracture was in every case that of the dotted line c, shown in 

 fig. 3. Professor Unwin, to whom this point was submitted, supposes 

 that this has no bearing on the strength or weakness of the j^articular 

 forms of thread used, but was due to internal strains in the steel produced 

 by the water-hardening, and to a slight bending force acting with maxi- 

 mum effect at the point where the screw enters the nut. The spherical 

 seats of the collars C C will not, he points out, wholly prevent the occur- 

 rence of this force. He suggests that had the screws been hardened in 

 oil this probably would not have happened. The sectional area of the 

 cores of the screws was "09.5 square inch, and the breaking strain was 

 expected to be about 1.3,. 500 lb. Those that broke where they entered 

 the nut broke at 5,600 lb. (60° screw), 5,860 lb. (50° screw), and 5,330 lb. 

 (40° screw) respectively. In a subsequent test one of them broke along 

 the line d (fig. 3) at 10,280 lb. Fortunately, in every case a sufficient 

 length of the screw was left after the accident to put on another nut, and 

 in each case a satisfactory result was obtained in a subsequent trial, the 

 screws being drawn thi'ough the nuts without being themselves broken. 

 The forms of the screws tested are shown in figs. 5 and 6. The 



Fig. 5. 



•5475^ 

 •AA/S 



diameters, both at the top and the bottom of the thread, were the same in 

 all the screws, and also in all the nuts. The screw threads were in all 

 cases flat-topped, with slightly rounded, but nearly flat, bottoms. The 

 pitch of the screw was the same in every case, 16 to the inch. The three 

 threads had angles respectively of 40°, 50°, and 60°. Each screw was cut 

 with a single point tool, ground to the correct angle from a cylinder pre- 

 viously turned to the correct diameter. The nuts were cut with single 

 point inside turning tools, also accurately ground, in a cylindrical hole 

 previously bored to the correct diameter. 



The outside diameter of each screw was -i\- inch (-4375) ; and the inside 

 diameter of each nut was "3575 inch. A clearance of "005 inch was 

 given at the top and bottom of the thread in every case. 



Fig. 5 shows the form of the screw and nut, having an angle of 40°, 

 and the dimensions which are figui^ed are the same in each of the other 

 two cases. 



Fig. 6 gives the outline of the contact surfaces of the screw and nut 

 in each of the three cases. In this figure the dimensions are figured in 

 thousandths of an inch. 



The dimenuions employed foi" these ecfetva were chosen as Rken, which, 



