The data recorded for each series of impact tests (whether hydro- 

 phone or accelerometer) were as follows: 



• The digitized impact waveform 



• The digitized response waveform received by the hydrophone or 

 the accelerometer 



• The location of the hammer impact and of the hydrophone or 

 accelerometer 



• The separation distance between the impact and the received 

 response 



• Description of the pile, including a sketch showing the damage 

 in test locations 



• Settings on the programmable oscilloscope 



• Time, date, and location of the test 



• Tape and file number that locates the stored waveform 



• Any changes in test setup or execution 



• Comments, including noticeable changes in recorded waveforms 



Laboratory Test Results 



The first series of hydrophone impact tests were carried out on the 

 standard pile with its simulated borer damage. These tests revealed 

 that the damaged portion of the pile had more discrete frequency bands 

 with smaller amplitudes than the solid portion, which had fewer narrow 

 frequency bands with higher amplitudes (Figure 24) . Repeated tests in 

 the same location resulted in similar waveforms and corresponding FFT's, 

 demonstrating the repeatability of the hydrophone impact tests. 



The accelerometer impact tests verified the hydrophone impact test 

 results; however, some discrepancies between repeated tests began to 

 develop. The accelerometer has a higher sensitivity and can detect 

 ambient noise to a greater extent than the hydrophone. Readings were 

 not always reproducible. 



The PCB hammer has an obvious effect on the impulse transmitted 

 through the timber pile and generates a transient (impact) excitation 

 signal. This is evidenced by a short narrow constant energy spike. A 

 single impulse is vital to initiating a wide range of frequencies at 

 essentially constant energy into the timber pile. To be sure that the 

 impulse generated by the PCB hammer was a single constant energy pulse, 

 the FFT of the input signal was analyzed. 



41 



