CONCLUSIONS 



The figures and graphs of this paper are 

 based on a laboratory study of the effect of air 

 gaps and saturated voids on neutron moisture 

 measurements. These results reflect the effect 

 of symmetrical voids, each 10 inches long. Sug- 

 gested procedures for estimating the amount 

 and type of error (positive or negative) in field 

 data may not yield accurate estimates for those 

 voids which are greater than 3 inches in radius 

 or shorter than 10 inches, or both. The Troxler 

 equipment used in this laboratory study is 

 nondirectional with respect to the axis of the 

 void; therefore, the use of other neutron equip- 

 ment which exhibits directional characteristics 

 may yield significant differences when field 

 studies are conducted on asymmetric voids. 



The study described in this paper provides 

 estimates of the maximum neutron measure- 

 ment errors which may be expected for sym- 

 metric voids in a sand medium when the water 



is at different levels relative to the center of the 

 void. Also, this study has shown that in the 

 field the presence of a void can be determined 

 most accurately when the soil is saturated. The 

 magnitude of the positive measurement error 

 v^dll exceed that of the negative measurement 

 error as long as any free water remains in the 

 void. 



Procedures which may be used to evaluate 

 the magnitude of the measurement error from 

 field data are outlined in this study. These 

 procedures may be used to correct data for 

 the effect of saturated voids surrounding the 

 access tube, but they in no way reduce the 

 importance of installing access tubes in such a 

 way that there will be no voids next to the 

 tube. Installation techniques together with de- 

 scriptions of the equipment used have been 

 reported elsewhere. 



19 



