13 



The recorder in its present form did the job well for which it was 

 designed. It has shown itself to be rugged and reliable with the exception of 

 two minor items, namely the relay failure mentioned above and a cam follower 

 which should be made of somewhat heavier steel strip. It would be a great con- 

 venience to incorporate into the recorder mechanism a device which would auto- 

 matically stamp the date and time on the record whenever a record is made. 

 Other desirable refinements would be a provision for a greater chart capacity, 

 say 200 feet, and two or three additional recording channels in order to per- 

 mit recording of strain, ship's speed, and heading simultaneously with the 

 other quantities. 



The mechanical design of the recorder is very compact and utilizes 

 subassembly construction, making for easy servicing. Many parts are standard 

 items, readily available. The cost of the recorder including the three accel- 

 erometer units was comparatively reasonable, but if additional channels were 

 added the cost may increase somewhat disproportionally because a standard 

 chassis might not be available. 



Analysis of the records obtained from the CASCO sea trials has been 

 made in terms of the number of cyclic variations of the magnitudes (roll, 

 pitch, etc.) that fell within given limits. Such an analysis can be presented 

 rather conveniently in the form of frequency distribution curves. This analy- 

 sis, although readily made by inexperienced personnel, is tedious; it could 

 be performed automatically by means of a counter system similar to the type 

 that has been under development for the strain cycle gage, see Reference 3- 

 Both the chart recording and the statistical analysis provided by the cycle- 

 counter data could be made available at the same time. The development of 

 this combination of instruments is highly desirable for the study of the 

 strength of ships at sea that is presently underway at the Taylor Model Basin. 



The following comments concerning the transducers are in order. 

 The individual accelerometers measure accelerations perpendicular to their 

 bases, and if the ship rolls, the heaving acceleration measured is the com- 

 ponent in the centerline plane of the ship. The heave accelerometer will also 

 be affected by the change in the component of gravity along its sensitive axis. 

 The error due to the latter cause is generally not serious except for the 

 larger angles of roll; for the largest measured rolling angle this effect 

 would be about 20 percent of the total signal. However, the heave at this 

 time was small. The proper correction can easily be made. Furthermore, these 

 effects are minimized by the fact that, when the heaving Is large, the roll is 

 generally small, and vice versa. In measuring the angular accelerations, the 

 effects due to the change in the gravity field or due to linear accelerations 

 are canceled out. 



