Adequate control of rock quality proved a difficult assignment. 

 Constant effort by both the contractor's quarry force and the engineer's 

 field staff was required in order to ensure a supply of rock of adequate 

 quality. Rock specifications required two quality tests -- the Los 

 Angeles rattler test, and the sodium sulfate soundness test. The sound- 

 ness test was considered especially important because of the marine expo- 

 sure. Rock quality varied widely throught the quarry site. Much of the 

 rock, of the Cold Water Sandstone Formation, Eocene Age, was of excellent 

 quality, but some deposits were poor and weakly cemented. There were also 

 many intermediate grades. The quarry site, however, contained a vast 

 amount of good material of a type which could be quarried in very large 

 unit sizes. 



During the initial quarry development the contractor drove many small 

 tunnels or coyote holes into the canyon sides searching for the best 

 quality rock. The coyote holes in sound rock were later used for primary 

 blasting. An extensive field testing program was carried out by the engi- 

 neer on samples taken from these coyote holes, and a search was made for 

 any quickly identifiable characteristics which would correlate with soimd- 

 ness. Acid reaction, specific gravity, Schmidt -hammer reading, color, 

 density, grain size, and microscopic examination were all tried, but found 

 unreliable. The final solution was to test each separately identifiable 

 rock type found in the quarry and to classify each according to its actual 

 soundness test results. Over 100 samples were required for adequate cover- 

 age. To speed testing, the engineer's field office was equipped to perform 

 soundness tests on a continuous basis. Untested portions of each sample 

 were retained and small chips from these were carried in a compartmented 

 box as an aid in quick field identification of the rocks. This method 

 proved effective in most of the cases, but a few types of rock, which 

 straddled the acceptance line, remained difficult to classify throughout 

 the job. 



Rock quantities were measured for payment by barge displacement. 

 Occasional checks of barge-gaging accuracy were made by weighing all loads 

 on truck scales and agreements were normally within 1 percent. Individual 

 weights of armor stone were normally judged by eye. In doubtful cases, 

 weight was checked by truck scales or by measuring the volume of the rock. 



Voids in the rock materials, as placed, varied from 40 percent for 

 armor rock which was nearly of uniform size to 30 percent for Class G 

 material of reasonably dense gradation. If losses of core fill material 

 are ignored, the tonnage placed would indicate about 20 percent voids in 

 the core, most of which was placed under water. A more reasonable assump- 

 tion of 35 percent voids in place indicates that about 23 percent of the 

 core material was lost due to ocean currents and wave action. 



A model of the island was built by the engineer's field staff at a 

 scale of 1:120,000. Progress on the model matched construction progress, 

 so that the model served as an easily understood progress report. While 

 the island construction was still below water, the model was especially 

 useful for visualizing the status of the work; and all persons connected 



18 



