sound velocity about 10% assuming the sound velocity through the damaged 

 concrete dropped to 8,000 ft/sec. If reference measurements could have 

 been made near the top of the piles, this change might have been detected. 

 As it stands, the average sound velocity through the 1930 piles appears 

 to indicate sound concrete when compared to similiar data from the 1958 

 piles that are not effected by the sulfate attack. 



Indirect ultrasonic measurements were made on the same piles as the 

 direct measurements and at essentially the same locations. It was assumed 

 that comparison of data from the two different pile groups would show 

 some indication of the sulfate attack occurring on the 1930 piles. The 

 data collected from the indirect measurements are tabulated in Table 10. 

 The table lists the pile, general location of the measurement, measured 

 pulse transit time over the fixed 10-inch path length, and general com- 

 ments concerning the shape of the waveform displayed on the oscilloscope. 

 Acoustic coupling was much more critical for these measurements compared 

 to the direct measurements because of the reduced signal levels at the 

 receive transducer. 



The average indirect transit time for the ultrasonic pulse in the 

 1930 piles was only about 1% higher than the average transit time for 

 the 1958 piles. If only the data from "good" waveforms are considered, 

 the difference is around 3.5%. From the direct sound velocity measurements, 

 a difference of about 4% would be expected in the indirect measurements, 

 which is the case if the data for the attenuated waveforms are dropped. 

 Dropping these data has the most effect on the average for the 1958 piles. 

 The 72-microsecond transit time for the top of pile 90H compared to 

 64 microseconds at the bottom corresponds to a half-wavelength difference 

 in path length at 54 kHz. Therefore, these data do not show any signifi- 

 cant differences in the ultrasonic pulse transit time through the 1930 

 piles compared to the 1958 piles. Thus, indirect ultrasonic measurements 

 did not detect the sulfate attack occurring on the 1930 piles. 



Indirect measurements were also taken on the pile caps in the same 

 location as the direct measurements. The indirect transit times were 

 quite large compared to the indirect measurements taken on the piles. 

 This can be illustrated by calculating an apparent sound velocity using 

 10 inches as the path length for the indirect measurements. For the 

 1930 piles this would be an average apparent sound velocity of 

 12,350 ft/sec, which can be compared to the direct measurement of 

 14,925 ft/sec, a decrease of about 17%. Calculating the same apparent 

 sound velocity for the indirect measurements on the pile caps results in 

 a value that is 45% lower than the direct measurement. This indicates 

 that a rating system could be developed and applied to indirect 

 measurements on concrete that has only one side accessible, even though 

 a path length is not very well defined. 



In summary, neither the direct or indirect sound velocity 

 measurements clearly indicated the sulfate attack on the piles in the 

 1930 group. The direct measurements indicated very sound concrete and 

 did not indicate the sulfate attack (comparing data from the 1930 and 

 1958 pile groups). Likewise, there was no significant difference in the 

 data from the two pile groups for the indirect measurements. Good 

 reference readings obtained above the waterline on each pile would have 

 provided more information. 



19 



