wood. The amplitude and strength of this portion of the signal is 

 affected by the internal condition of the timber pile. The part of the 

 signal traveling through the water indicated by a "2" in the figure 

 arrives later but with a larger amplitude. 



The fixed distance between the transducers was set at 19 inches 

 (0.483 meter) for laboratory tests. The travel time for the sonic 

 signal to travel directly through the timber pile from the transmitter 

 over the fixed distance to the receiver would be equal to 97 usee 

 (0.483 m t 5,000 m/sec) . Consequently, the first time frame of 100 usee 

 relies heavily on the arrival of only the ultrasonic signals traveling 

 through sound wood. The travel time for the sonic signal to travel 

 through a damaged pile section is influenced by the amount of water con- 

 tained in the damaged area. If the damaged area were completely filled 

 with water (e.g., from borer damage), then the time for the sonic signal 

 to travel from the transmitter to the receiver would be 320 usee 

 (0.483 m t 1,500 m/sec). Therefore, the ultrasonic signal contained in 

 the 400 usee bin is influenced mainly by the slower, larger amplitude 

 waves traveling through water. The 200- and 300-usec bins are inter- 

 mediate time frames that include the waves traveling partially through 

 water-filled voids and solid wood along with the stronger amplitude 

 waves traveling only through water. 



The location of the transducers was recorded using the quadrant 

 number (Q) and the transmitter-to-receiver increment position (TX-RX) . 

 A data entry identified by Ql-2 RX1 TX8 means the ultrasonic signal was 

 transmitted through the wood between quadrants 1 and 2. with the trans- 

 mitter at the 8th marked line and the receiver at the 1st marked line. 

 The distance between the marked lines varied for each test pile as 

 explained previously. 



After completing the laboratory ultrasonic and impact tests, the 

 five laboratory test piles were sectioned and photographed. The photo- 

 graphs of the various sections for each pile were enlarged and the cross- 

 sectional area was divided into quadrants. An example is shown in 

 Figure 12. With the photographs and an engineering tool called a planim- 

 eter, the cross-sectional area remaining in each quadrant was calculated. 

 Subtracting the remaining cross-sectional area from the total cross- 

 sectional area gave the amount of cross-sectional area loss. Correlation 

 between the amount of cross-sectional area loss in each quadrant and the 

 standard deviation of the ultrasonic signal received in the respective 

 quadrant was investigated. 



Figure 13 shows the relationship between the cross-sectional wood 

 loss of the standard pile and the standard deviation over four different 

 bins of the ultrasonic signal. The best correlation between the standard 

 deviation and the cross-sectional wood loss was found to be over the 

 100-usec time bin (Figure 13). Standard deviation generally decreases 

 with an increase in the percent of cross-sectional wood loss. It is 

 apparent from the figure that the data collected contain discrepancies. 

 Standard deviation (over 100 usee) for a 17% cross-sectional wood loss 

 is higher than standard deviation where a 9% cross-sectional wood loss 

 exists. Repeated readings often varied in the same location. Signal 

 variations that occur through a constant amount of cross-sectional loss 

 are due to material changes in the structure and grain orientation of 



25 



