The MCLS was again returned to the contractor. 

 Although the evidence is not completely definitive, 

 the most probable cause for the traction-winch drive 

 failures in 1973, as well as in 1972, appears to be 

 high-intensity impact loadings occurring when the 

 boom hits the bottom stops and the payload is 

 abruptly accelerated upward. This process creates 

 very large dynamic-tension loads in the lift wire that 

 are transmitted directly to the traction-winch drums. 

 Because of the extremely short rise times associated 

 with these tension pulses, the loading on the pinion 

 gear, which drives both traction-winch drums, may 

 not have been symmetrical. Nonsymmetrical forces 

 on the pinion would, in turn, create large side loads 

 on the pinion shaft and bearings, much larger than 

 they were designed to withstand. In fact, examination 

 of both drive systems does show massive failure of 

 the pinion drive-shaft bearings and associated bearing 

 retainers. 



TEST RESULTS 



In spite of the many problems encountered 

 during the at-sea test series, useful data were obtained 

 on the performance of the MCLS. The two basic 

 parameters of interest are the dynamic tension con- 

 trol and the payload motion. Due to a variety of 

 instrumentation difficulties, valid payload motion 

 data were not obtained. Comparison of boom-tip and 

 deck vertical-acceleration records indicates that some 

 motion reduction was obtained, although the magni- 

 tude of the reduction could not be determined. 



Valid dynamic tension data were obtained for all 

 tests except the one conducted on 6 October 1973. 

 Tension data are considered valid for intervals (data 

 sets) in which no boom-impact-induced dynamic 

 loading occurs. In most cases valid data in both hard 

 and soft modes exist for each test so that the reduc- 

 tion in dynamic tensions attributable to the MCLS 

 can be determined along with the dynamic/static 

 tension ratios. For each test record examined, a series 

 of data sets representative of the entire record were 

 chosen; each set was a tension-time record of 30 to 

 60-second duration. For each set, the envelope of 

 peak dynamic tension variations was determined 

 along with the mean, or static, tension. Table 2, a 

 summary of the dynamic tension data, shows that the 

 dynamic-line tension was, on the average, controlled 



to less than 14% of the static value. This is a 

 reduction of about 2 to 1 in dynamic-line tensions 

 over themoncompensated mode of operation. 



SUMMARY 



In considering the potential for motion- 

 compensating lift systems of the passive, fluid-spring, 

 boom-bobber type described in this report, it is 

 necessary to differentiate between the limitations of 

 the basic concept and the hmitations of the specific 

 hardware. It is believed that a boom-bobber system 

 using a passive fluid spring is capable of providing the 

 motion compensation and line-tension control to the 

 degree specified in the original Request for Proposal 

 (RFP). However, careful attention must be given in 

 the designing process to the type of problems encoun- 

 tered with the existing MCLS hardware as described 

 in this report. 



In a review of the past series of tests, a number 

 of problem areas that contributed directly to the 

 system failures become evident. One of the most 

 critical problems seemed to be the lack of under- 

 standing of the importance of the system's non- 

 linearities and the need for a comprehensive 

 quantitative analysis of all proposed hardware prior 

 to fabrication. Parameters such as the boom-spring 

 geometry, variable gas "constant," internal "stick- 

 tion,"* internal damping due to fluid flow, and 

 control-system characteristics must be included in the 

 analysis. At the time of the present system's 

 development it was felt by the contractor that a 

 linear approximation would provide a valid represen- 

 tation of system performance. It is now apparent that 

 only a comprehensive nonlinear computer model of 

 the system, used with random ship-motion data for 

 input, is adequate to provide realistic modeling of 

 expected system performance. 



Another questionable area was what happened 

 when the boom tried to travel beyond its normal 

 stroke limitations." Again, hindsight shows that some 

 sorl of progressive hardening, or cushioning, of the 

 boom is required as it approaches either end of its 

 stroke. This feature is mandatory to prevent the 

 damaging impact loading experienced with the 

 existing MCLS when the boom hit the stops. It is 

 believed that any seagoing system such as the MCLS 

 will, sooner or later, be pushed to or beyond its 



* System breakaway friction. 



24 



