tests, the density ratio was pyp^a =1.81. Therefore, the method of Thompson and 

 Shuttler yielded a damage index approximately nine times that of Broderick and Ahrens, 

 or the width of their structure in nominal diameters. Thompson and Shuttler also 

 determined the minimum armor layer thickness at failure. They defined failure as the 

 point at which an area of exposed underlayer of diameter D„sq occurred. Note that the 

 minimum armor layer thickness will not be zero at failure because it is expressed as a 

 spatial average of several profiles. 



H. R. Wallingford, Ltd. (1990), showed that Equation 3.1 yielded very 

 different results if a slightly different method was used to compute the average eroded 

 area. The first method they used was that described for the WES eroded volume 

 method, where an average profile was used to determine an average eroded area. The 

 alternative method was to sum the eroded areas from all profiles in order to compute an 

 average eroded area. The difference between the two methods ranged from 2 to 82 

 percent. In general, the difference decreased as the damage level increased. Note that 

 most authors do not describe the method used to compute damage. 



All damage methods discussed above share a common weakness, namely 

 they compute the average damage, which may be concentrated in one pocket or spread 

 out over several areas. Also, none of the methods give any indication of the profile 

 shape, or more specifically, the maximum depth of erosion, which is certainly an 

 important parameter for a multilayer structure. 



39 



