172 COLLAPSE OF TEXAS TOWER NO. 4. 



mately 9 inches; its lockiug devices and keeper pin missing. All other pins, 

 connections, and bracings were found to be in excellent condition at the minus- 

 23-foot level. 



During the repairs, a temporary locking device was fabricated and was 

 installed to insure the pin's staying in its present position. The pin, although 

 withdrawn 9 inches, was still in shear with all parts of the connection. 



B caisson collar at the minus-65-foot level was cleared of marine growth. The 

 collar was found to be without vertical motion ; all bolts, nuts, and double nuts 

 were found to be in place. However, the Dardelet keeper plates were loose and 

 several of the studs and nuts were missing in the keeper plates. A rectangular 

 plate three-eighths of an inch thick was installed over the heads of the Dardelet 

 bolts to stay iu position. The plate is secured to the caisson by means of eight 

 ^,^-inch studs welded to the caisson. This plate was installed in the field and 

 does not show on any construction print. The collar holding bolts were found 

 to be at an average of 900-foot-pounds torque. These bolts were torqued to 1,300 

 foot-pounds. They were foimd to be holding at 1,300 foot-pounds after slight 

 sea conditions. 



As well as could be ascertained without removing the keeper plates of the 

 Dardelet bolts, we found that several of them were loose at the upper and lower 

 extremities of the collar. 



On further examination of B caisson, we found two double sheave blocks and 70 

 feet of wire cable hanging from shackles at the minus-70-foot level. These 

 blocks were removed and salvaged, the cable dropped. This removal was made 

 at the request of the naval inspector, Mr. Kelleher. 



A caisson collar at the minus-65-foot level was cleaned of marine growth over 

 its entire connection area. We found the caisson in the way of the top and 

 l>ottom edges of the collar shined to bright bare metal. This bright area, approx- 

 imately two inches in width, was caused by vertical motion in the collar. Upon 

 further examination, we found that several of the jam nuts and double nuts 

 that secured the 2-inch collar hold bolts had slackened away from the holding 

 nut an inch or two. It was also found that the Dardelet bolt keeper plates 

 were missing. 



None of the Dardelet bolts on either side were in place, having either fallen 

 out or sheared off. We found six Dardelet bolt shanks still in the caisson. 

 The bolt holes had the appearance that the bolts had been i-lieareil by the 

 vertical motion of the collar ; that is, the holes had a top and bottom elongation 

 rather than a true round hole. Each of the holes with the elongation seemed 

 to have been radiused on the extreme edges, indicating that the bolts had finally 

 sheared after considerale vertical motion. 



Concerning the collar holding bolts and nuts, it was found that several of the 

 nuts were hand tight. Our first operation endeavoring to torque the bolts to 

 1,300 foot-pounds gave us no indication of torque on any of the bolts. However, 

 we were using a 2,000-pound meter on an extremely large wrench such that 

 there may have been slight torgue on some of the bolts not indicated by the 

 heavy wrench. We tightened all bolts with a hand wrench and then torqued 

 to 1,300 foot-pounds. 



Over a period of 1 night, we found on examination the following day that 

 the bolts were again out of torque. We had found on close exanxination at 

 the top and bottom edges of the collar that the collar was not conforming 

 100 percent to the caisson. Consequently, when the bolts were torqued to 

 1,300 foot-pounds, the collar was drawn closer to the caisson so that our 

 1,300 foot-pounds was not true. The reason being the collar could not be 

 drawn to full friction with the caisson in one operation of tightening. Through 

 the night with the working of the collar and bracing, the collar would further 

 conform to the caisson. We then torqued the holding bolts up to an average 

 of 1,800 foot-pounds endeavoring to draw the collar into friction in one 

 operation. 



This was successful in stopping the vertical motion. However, the torque 

 over a period of 2 standby days was reduced to 1.300 foot-pounds. We again 

 torqued to an average of 1,800 foot-pounds and in this operation we were suc- 

 cessful in holding the bolts at a constant torque. The collar at this time was 

 bearing in friction approximately 80 percent. 



We checked the pin at the extremity of the diagonal brace from the minus- 

 65-foot level on A caisson to the center of the minus-23-foot horizontal span 

 from A to B caisson ; this is the pin that was found to be out of shear and 

 loose in the initial survey. At this time, this pin was in full shear. The 

 jamming nuts were brought to bear for the last time during this repair 

 period. 



