22 COLLAPSE OF TEXAS TOWER NO. 4 
office in Bath, Maine, had found it necessary to approve an increase 
in tolerance in certain instances greater than those initially called 
for in the Moran-Proctor structural drawings because of difficulties 
encountered in fabricating the template. 
During the winter of 1958-59, winds to 65 knots and waves to 30 
feet were experienced over a 5-month survey. From the Brewer 
study it was found: 
(a) That the observed natural tower frequencies (17 to 23 c.p.m. 
translational and 23 to 24 ¢.p.m. rotational) were approximately one- 
third of those predicted by the designer’s theoretical calculations. 
(6) That the subsea truss work was essentially ineffective for ex- 
cursions up to 3 inches and rotations to 0.1°. It was expected that the 
clearances in the pin connections would be taken up with increasing 
deflections of the tower platform. 
(c) That positive evidence of the fact that relative motion between 
members of the underwater truss system occurred during the ever- 
present tower oscillations, was provided by hydrophones. 'The metal- 
lic rumbling noises heard beneath the tower were coincident with the 
frequency of tower motion. They were interpreted to result from 
the movement of very heavy metal objects. 
(d) That 10-foot waves produced the greatest tower motions and 
therefore stresses over the range of waves 0 to 30 feet in height ex- 
perienced during the study. 
(e) That hydrodynamic forces (waves) were by far the more im- 
portant over aerodynamic forces (wind). 
It was recommended by Mr. Brewer that, if the extent of the plat- 
form excursions and rotations were objectionable, then an investiga- 
tion should be conducted to determine whether greater rigidity might 
be achieved by installing bracing above water to reduce the bending 
moments of the legs. This suggestion was limited to a means by 
which to reduce platform motion and was not intended as a means of 
strengthening the tower. Mr. Brewer testified that a complete stress 
reanalysis would have to be made because such bracing would increase 
the resistance to wave passage and, therefore, could conceivably have 
the effect of actually weakening the tower. 
From merely a cursory computation which was not required as a 
portion of his responsibility under his contract, Mr. Brewer found 
that if the tower legs were to have no bracing, the weight of the plat- 
form alone would collapse the legs without any wind or wave force 
being exerted against the tower. 
The design engineers, however, found fault with the Brewer study 
by the following comment: 
During the fall of 1958, a subcontract was given to the Brewer Engineering 
Laboratories, Ine., to perform motion studies on the tower. These consisted 
of horizontal acceleration measurements in the vicinity of each of the three 
legs correlated with strain gage measurements on the legs and simultaneous 
observations of wind and wave direction and amplitude. These studies were 
not very successful mostly because it was very difficult to determine the true 
translations which involved the multiple integration of the curves obtained 
from rather irregular data.” 
There is nothing mystic or mysterious about multiple integrations 
to determine displacements. Although done automatically by com- 
17 ‘Design and Construction Report,” p. 64. 
