144 COLLAPSE OF TEXAS TOWER NO. 4 



( 6 ) An 87-mile-per-hour wind with a 67-foot wave. 



The tower was designed for a 60-foot wave accompanied by a 125-mile-per- 

 hour wind while at the same time the maximiun allowable stress was permitted 

 to increase one-third above the basic stress. 



It would seem entirely possible to get a 125-mile-per-hour wind as this is 

 not rare in tornadoes or hurricanes and it must be assumed that the maximum 

 wave either preceded, or lagged behind, the maximum wave gust, which is a 

 more or less likely supposition. 



In any case it is of considerable concern to think of the tower having to sustain 

 the impact of a hurricane without the upper bracing in place. We still recom- 

 mend a diver inspection of this tower with special attention being given to the 

 bolted collars which had to be substituted for the destroyed connections at the 

 lower ends of the upper diagonals on the A-B side. Since these diver programs 

 are diflScult and costly to organize, the diver should probably go out with equip- 

 ment suitable for tightening the collars if they should be found to be loose. 



With regard to the often discussed program for measuring deflections of this 

 and other Texas towers, it can be said that information on the motion is desira- 

 ble from the standpoint of Air Force operations and for safety reasons. Actually, 

 the type of information which is needed for the latter probably differs quite 

 appreciably from that required for technical considerations involving signal 

 reception. A third purpose for recording motion would be a purely scientific 

 one involving correlation between wave and wind data, tower motion, and tower 

 stresses. A program to fit this third objective would be extremely comprehensive 

 and very diflBcult to sustain, due mostly to personnel problems associated with the 

 relative remoteness of the towers. The instrumentation, particularly the wave- 

 recording instruments, would be diflBcult to maintain. It is possible, however, 

 that, with the assistance of the Hydrographic Ofiice, some sort of limited pro- 

 gram might be set iip which would result in wave force information. It should 

 be mentioned that equivalent ieformation might be more directly and accurately 

 obtained if the complexities of tower strain and motion were not involved. This 

 sort of investigation is, however, a subject aside from measurement of tower 

 motion and accompanying strain. 



For indications of structural distress in a tower during severe meteorological 

 conditions a minimum instrumentation should provide for measurement of 

 tower deflection over separated periods and limited to times when tower motion 

 might be thought alarming or for other special reasons. Measurements of 

 translation in two directions would be made at a location as near as possible 

 to the geometric center between the legs and no attempt made to measure 

 rotation. This would resemble in purpose the program undertaken on TT-2 

 by the Lincoln Laboratories in 1956 but would be set up with a somewhat dif- 

 ferent approach. Whereas, they were concerned with small movements, and 

 also rotation, the new purpose would be to measure larger general movements of 

 translation only. Their approach to the problem was to correlate readings of 

 strain gages installed on the legs with seismic accelerometer readings so that 

 the strain gage readings could, at any subsequent time, be used to estimate the 

 motion. This method was not too successful mostly because the strains were 

 too small during the calibration periods and too small, in any case, to measure 

 the motions in which the sponsoring parties were interested. Also, the strain 

 gages and associated lead wires were in extremely uncomfortable and diflScult 

 locations in the legs which greatly discouraged if not prevented maintenance. 

 The minimum program offered here would be to have accelerometers permanently 

 in place to measure motion in two directions only. No attempt would be made 

 to simultaneously measure strains (and stresses) but rather these would be 

 computed from the recorded deflections. 



We investigated the instrumentation problem in 1956 and, after considerable 

 diflBculty in locating interested parties, secured a proposal from Arde Associ- 

 ates of Newark, N..I., for assembling and installing a system involving two accel- 

 erometers at each corner (a total of six) which, with electronic integrating ele- 

 ments, could be used to register translation and rotation. With the program 

 suggested hereinbefore only two such instruments would be used, which would 

 measure only general translation and no rotation. 



The Arde proposal was sent to your oflRce in May or .Tune 1956 and we re- 

 ceived a reply from the ofl3cer in charge of construction that the study was not 

 required "at this time." You will find in the Arde Associates' proposal an 



