2.286 m 



o 

 CO 



Figure 1. 



A-METER 

 B- STEEL CABLE 

 C-BRASS ROD 

 D- SHOCK CORD 



-Diagrammatic view of 3-m IKMT mouth with 

 mounted flowmeter. 



and net removed, was suspended in the IKMT 

 mouth. This sampler had been calibrated by 

 carefully controlled tows over distances of 2 and 

 6 nautical miles before being installed in the 

 IKMT. Start and end positions for each tow were 

 determined by radar, LORAN, and shore fixes. 



The three legs of shock cord used in the pre- 

 vious trial were attached directly to the frame of 

 the C-B sampler so that the flowmeter could 

 pivot freely within its frame. During launch and 

 again as soon as the flowmeter cleared the water 

 during retrieval, the flowmeter automatically 

 pivoted, causing the axis of the impeller to lie 

 perpendicular to the direction of the tow; i.e., the 

 meter did not register. The impeller blade housing 

 protected the blades from winds, thus preventing 

 rapid spinning of the impeller and erroneous flow 

 readings. As soon as the net was lowered and in 

 towing position, proper aspect of the impeller axis 

 (parallel to the flow) was maintained by water 

 pressure acting on the stabilizing fins attached to 

 the impeller blade housing. 



Two net tows were carried out using this 

 apparatus, with the shock cord again stretching to 

 place the C-B sampler just inside the net 



mouth. Start and end positions were obtained, as 

 during the calibration runs, by radar, LORAN, 

 and shore fixes. Previously determined calibra- 

 tion information allowed calculation of the 

 number of revolutions that would have resulted 

 if the sampler had been towed by itself over the 

 same known distances covered during the two 

 tows. 



All net tows were made at ship speeds ranging 

 between 3 and 4.5 knots over distances of 6 and 

 10.75 nautical miles. The direction of the tows was 

 approximately perpendicular to the prevailing 

 current flow. Very little biological material was 

 captured during these net tests, hence clogging 

 was considered a negligible factor. 



Results and Discussion 



The two methods used to determine filtering 

 efficiency yielded similar results (Figure 2). For 

 method 1, the total number of revolutions (N') 

 registered by the meter in the net mouth is 

 plotted against the number of revolutions (N) 

 registered by the meter on the spreader bar for the 

 two net lowerings. These results are shown as 

 squares. For method 2, the number of revolu- 

 tions (N') registered by the calibrated C-B meter 



n INDICATES NUMBER OF REVOLUTIONS REGISTERED BY 

 THE METER ON THE SPREADER BAR (N) PLOHED AGAINST 

 NUMBER OF REVOLUTIONS REGISTERED BY THE METER IN 

 THE NET MOUTH (N'l FOR TWO NET LOWERINGS (METHOD 1). 



A INDICATES NUMBER OF REVOLUTIONS EQUIVALENT TO 

 THE DISTANCE TOWED (N) PLOTTED AGAINST NUMBER OF 

 REVOLUTIONS REGISTERED BY CALIBRATED C-B IN THE 

 NET MOUTH (N') (METHOD 2). 



100 



CO 



O 



X 80 



CO 



Z 



g 

 5 60 



6 

 > 



EFFICIENCY=92% ^t-. 



120 



N (REVOLUTIONS) x lO" 

 Figure 2.— Filtering efficiency of 3-m IKMT. 



620 



