The Value of Engineering Inputs to Analyses of Acoustic Data
By Bruce Rule - Nov 23, 2014
Russ: Please archive this posting..
As previously discussed, collateral assessments made primarily from System data have been of great value to the US submarine design community. For example, the requirement for a 30+ knot speed capabilty of the 688 Class SSN was, in part, a response to acoustic-derived assessments of the speed capabilities of the VICTOR Class Soviet SSN.
Early “non-acoustic” assessments of the speed capabilities of Soviet nuclear submarines were based primarily on the assumption that Soviet “packaging density” (horsepower per cubic foot of engineering space) could not exceed US values, e.g., NAUTILUS.
Based on that assumption, the 28-knot NOVEMBER Class was initally assessed to be capable of only 22-knots, the 45-knot PAPA only 28 knots, and the 41-knot ALFA also 28-knots. When one looks at early estimates of installed horsepower, we have 15K for the NOVEMBER versus the Russian published value of 37.5K, and similarly low estimates for PAPA and ALFA when the subsequently published values were 80K and 40K respectively. It took acoustic data to establish that the packaging density estimates were seriously in error.
Making assessments of design characteristics from acoustic data can be tricky business requiring information not normally available to those activities involved in such derivations. The recently discussed assessment of the use of a reduction gear by Project 636 KILO Class Russian submarines is such an example.
The writer used detailed open-source schematics of the internal arrangements of Project 877 and 636 KILO Class submarines to derive an estimate of 380 cubic feet for the PG-141 and PG-141M main propulsion motors which have maximum speeds, respectively, of 500 and 250 RPM. The writer then went to the Electrical Engineering Department of the University of Kentucky for estimates of the volume of dc motors of the same horsepower developed at those rpm values. The answer: twice the volume for half the rpm required that the PG-141M motor use a built-in reduction gear – as previously discussed. Bottom line: a dc motor that developed 5,550 hp at 250 rpm would be too large to install in the available space within Project 636 hulls.
A second – and similar - example involved an assessment that the Soviets had developed a full-speed-range turbo-electric propulsion system for a series production submarine. Same result: such an electric propulsion motor would have been too large and too heavy to install in the candidate hull. The answer in that specific case was a reduction gear which was both space and weight efficient compared to an electric drive of that vintage capable of achieving the demonstrated speed.
A third example involves Soviet/Russian GTZA (main propulsion gear assembly) nuclear submarine propulsion components in which the turbine and the reduction gear are essentially a single unit mounted on the same sound-isolation “raft.” Such a design is consistent with the turbine and reduction gear acoustic components having similar acoustic detectability characteristics. If one component of such a GTZA is significantly more detectable than the other, then the source assignments come into question, i.e., one is probably wrong.
There may – repeat may – still be a major position extant that violates this no “acoustic dichotomy” axiom. In this case, the basic value, refined to the fourth decimal place is precisely the quotient of two integers, the (much) larger of which is prime, a characteristic suggesting a single-stage reduction gear with a high ratio. This assessment is anther example of the most accurately you measure something, the more you find out about associated areas.
As long as acoustic analysis remains a primary source of otherwise unavailable information on foreign submarine design and operational characteristics, it should be the responsibility of those who derive such estimates to, as previously stated, “think beyond the gram” and seek useful collateral information wherever it can be obtained, especially from public sources including open literature, i.e., don't analyze acoustic data in a collateral information vacuum.