MARITIME ENGINEERING JOURNAL Maritime ^, Engineering Journal

National Defense 1*1 Defence rationale Maritime Engineering

Journal January 1989

•trt The Saguenay Gearbox |]Y> Mystery

Canada A look back at HMS Charybdis -page 29

MARITIME ENGINEERING JOURNAL Maritime ^, Engineering Journal

JANUARY 1989

DEPARTMENTS Editor's Notes 2 Commodore's Corner 3

FEATURES The Saguenay Gearbox Mystery by Lt(N) Kevin Woodhouse 4 CANT ASS — Bringing ASW into the 21st Century by LCdr Richard Marchand 9

Director-General Development of a Reverse Osmosis Desalination Maritime Engineering and System for the Naval Environment Maintenance by Morris Shak and Real Thibault 15 Commodore W.J. Broughton The "Black Art" of Propulsion System Alignment by LCdr Brian Staples 22

Editor Evolution of the Man/Machine Boundary Capt(N) Dent Harrison in Combat Systems by Cdr Roger Cyr 26 Technical Editors LCdr P.J. Lenk (Combat Systems) Cdr Roger Cyr (Combat Systems) LCdr Richard Svlvestre (Marine Svstems) LOOKING BACK: HMS Charybdis 1880-1882 29 Steve Dauphinee (Marine Systems) LCdr Richard B. Houseman (Naval Architecture) NEWS BRIEFS 30 Production Editor LCdr(R) Brian McCullough Index to 1988 Articles .. .. 32 Graphic Design Ivor Pontiroli DDDS 7-2

Word Processing by DMAS/WPRC 4MS Mrs. Terry Brown, Supervisor

The Maritime Engineering Journal (ISSN 0713-0058) is an authorized, unofficial publication of the maritime engineers of the Canadian Forces, published three times a year by the Director-General Maritime Engineer- ing and Maintenance. Views expressed are those of the writers and do not necessarily reflect official opinion or policy. Correspondence can be addressed to: The Editor. Maritime Engineering Journal, DMEE. National Defence Headquarters. MGen George R. Pearkes Building. Ottawa. Ontario, Canada K1A OK2. The editor reserves the right to reject or edit any editorial material, and while every effort is made to return artwork OUR COVER and photos in good condition the Journal can assume no responsibility for this. Unless otherwise stated. HMCS Saguenay Journal articles may be reprinted with proper credit.

JANUARY 1 1 Editor's Notes Due to circumstances beyond our control

Two years ago when we put the Unfortunately, the "price" for our So despite some of the frustrating set- machinery in gear to start the Journal on September issue ran considerably higher backs we have experienced in bringing its way towards a bilingual format, we and in a sense we are still licking our you a bilingual branch journal, we remain knew that the path would not be without wounds from it. Extraordinary delays of optimistic. The results, we feel, have by its pitfalls. And how right we were. up to three months in translation meant far outweighed the effort. What you see that we wouldn't be able to get a bilingual here is not grudging or token compliance Apart from our trial by fire in coming issue out until the beginning of December. with the Government's policy on bilin- up to speed on the ins and outs of editing It was a bitter pill to have to swallow, but gualism, rather it is a serious response to and producing a bilingual publication, we having set the new course back in April it. Today in the Canadian Forces it is the found that we had to contend with a seri- we felt there was no turning back. unilingual branch journal that stands out ously overtaxed translation service. as the anachronism. And that is something Delays in translation meant delays in get- By contrast, the translation of the issue which we can do well without. ting the manuscripts to typesetting — and you are reading now could not have gone on it went. But we were determined to more smoothly. Quick turnarounds and have a two-language branch journal excellent translations — we couldn't have before the end of 1988. hoped for better, and we'll never ask for less. Whether or not such service can be In the end we succeeded. With publica- maintained remains to be seen, but in the tion of our April 88 issue we were able to meanwhile we can applaud the steps that deliver the Journal to you for the first time have been taken in recent months to in a fully bilingual format. We realized accommodate the needs of the Journal. midway in the production of that issue that translation delays would make an April Among the initiatives being taken to distribution impossible, but a six-week streamline the bilingual production proc- delay seemed a small price to pay for ess, we have begun using desktop publish- finally getting our new format on the rails. ing to incorporate last-minute news items and more up-to-date commentaries. It becomes a bit of a tightrope act, but there is just enough time to get the late bits translated and set by hand before the major articles come back from commercial typesetting ready for paste-up.

MARITIME ENGINEERING JOURNAL OBJECTIVES • To promote professionalism among maritime engineers and technicians. • To provide an open forum where topics of interest to the maritime engineering community can be presented and discussed even if they may be controversial. • To present practical maritime engineering articles. • To present historical perspectives on current programs, situations and events. • To provide announcements of programs concerning maritime engineering personnel. • To provide personnel news not covered by official publications.

MARIT1MI-: ENGINEERING JOURNAL Commodore's Corner TASK from a Naval Perspective By Vice-Admiral W.B. Hotsenpiller

I am most grateful to your editor for an ship and/or occupational qualification, he The potential impact of TASK on naval opportunity to provide my final contribu- or she will be provided with tangible establishments, training, occupation tion to the Journal. Given the number of rewards, i.e. more pay, more authority structures and career management must years I have spent in the personnel world, and responsibility, and more challenging not be underestimated. Moreover, we can it should not come as a surprise that I have jobs. No longer will the navy be required and will get it right. In the final analysis, chosen to comment on a major personnel to conform to a structure where occupa- TASK should provide the navy with issue. Lateral skill progression, or TASK tional skills are rigidly linked to specific increased flexibility to: (Trade Advancement for Skills and rank levels and where all tangible rewards a. restructure our MOCs in response Knowledge) as it is now called, is are tied exclusively to rank progression. to new operational requirements; "older" than MORPS and if implemented The TASK structure would allow the mix- will have an even more profound effect on ing and matching of leadership and b. provide career progression pat- our sailors and their profession than occupational expertise and provide terns that can be tailored to meet our MORPS has had. rewards to meet the unique needs of each future technological and human needs; MOC and its serving members. Under c. provide a means to attract and As you are aware, the navy is currently this system, individuals could achieve retain skilled recruits without having to in the throes of a lengthy, complex and higher levels of occupational or technical confer rank upon them artificially; and challenging transition to a new CPF and expertise and be paid/rewarded for this finally, submarine fleet. This process will gener- accomplishment, regardless of whether ate many changes in the way we do busi- d. adjust our reward system, which they are promoted in rank. In turn, rank ness. MORPS, the most recent major includes promotion, pay, recognition progression would more clearly focus on personnel structure review, resulted in a (uniforms, badges and medals), the requirements to provide effective realigment of our NCM occupations and benefits and privileges to reinforce our leaders, supervisors and managers. capabilities to meet future fleet require- naval values and beliefs and meet the ments. The MORPS decisions were made How this can be achieved and the needs of our sailors. several years ago, based on our best esti- specific impact TASK changes would I am confident that the best interests of mate of what the new fleet would need. have on the navy and its sailors has yet to the navy and its sailors will be served by Since then our requirements have been be determined. Be assured however that this concept and I only wish that I was refined and new technological challenges this major undertaking is not an NDHQ going to be around to participate in its continue to be identified. As a result, fur- solution to an outdated CF problem. introduction. ther changes are in the wind. TASK is being developed in close cooper- To meet the navy's evolutionary opera- ation with Maritime Command and the tional, equipment and human demands, naval branch advisers who are collec- Vice-Admiral Hotsenpiller is the Assistant we require more flexible occupation and tively looking to the future. Although Deputy Minister (Personnel) career progression structures. To this PMO TASK is designing the broad policy end, PMO TASK has been given the man- framework, it will be the navy that date to develop a new rank, occupation decides how its MOCs will be structured progression and reward policy framework and rewarded within the framework by that will meet the future needs of the CF. means of a process which, among other features, will involve carefully chosen TASK is predicated on the need to representatives of each occupation. PMO quantify and reward leadership (rank) and TASK will also canvass approximately 10 occupation skills separately. This means percent of our sailors on both coasts to that as a sailor achieves a higher leader- determine their attitudes towards our current career progression and reward structure. This feedback will be acquired through a formal questionnaire to be given in mid-January 89. The Saguenay Gearbox Mystery

By Lt(N) Kevin Woodhome

A report on the investigation and Saguenay locked her port shaft and union to devise an investigation proce- repairs made to HMCS Saguenay's port slipped quietly into Rosyth for a closer dure. Time was of the essence because propeller shaft and gearbox. look. Saguenay was due to arrive in Halifax on December 5th, with docking scheduled Setting a Riddle During the ensuing week a mixed for the 12th. There would be only one progression of transatlantic telephone week to find out what was wrong with the HMCS Saguenay was forced to return calls served only to deepen the mystery. to Canada from Europe in November ship while she was afloat, before instruc- There was no observable damage at either tions had to be passed to the commercial 1986 after colliding with a West German port or starboard tailshafts or A-brackets, docking company which would be effect- submarine. During the incident, which yet the thrust shaft at the gearbox was ing repairs. A plan of action was finally occurred in August, minor structural found to be eccentric at the output flange agreed upon and set into motion. Permis- damage was sustained on the starboard by 0.030 inches. The only other abnor- sion had already been obtained from Can- side of the ship, abaft the engine-room mality, which had been reported earlier, ada Customs and Excise to commence and below the waterline, when the subma- was the poker gauge readings taken at the work on the ship even before she cleared rine passed underneath the ship diagonally port A-bracket. These indicated a shaft customs on arrival at Halifax. astern, eventually striking Saguenay's run-out of some 0.050 inches, but clear- port propeller. Two propeller blades ance within the A-bracket was known to Figure I shows the layout of received minor damage, one having two be in excess of 0.125 inches anyway; instrumentation which was to be attached or three inches of the tip removed, and the moreover, underwater poker gauge read- to the shaftline. The search for abnormali- other being slightly bent over at the tip. ings have a special notoriety of their own. ties along the shaftline was conducted by Apart from an increase in cavitation no measuring eccentricities at key positions. It was well documented that other ships vibration was reported and no abnormali- The insert shows how dial micrometers of the class had hitherto received signifi- ties were observed along the shaftline or were fixed to look for distortion in three cantly greater damage to propellers gearing. Whilst at Wilhemshaven the dimensions whilst rotating the propeller minor structural damage was repaired and (including having substantial portions of shaft. the propeller blades were ground fair blades severed by underwater grounding), underwater. without affecting propeller shafts in any For the uninitiated, bearing reaction way. This knowledge was later to prove testing is a rather clever method of finding By the end of August Saguenay was most misleading. out if the propeller shaft is bent at a given back at sea completing her STANAV- position. It was planned to check the reac- FORLANT deployment. All appeared The investigations at Rosyth were tions at the thrust block, plummer block quite inconclusive. The port gearbox out- well, and the staff of the Naval Engineer- and stern tube. The idea is to lift the shaft ing Unit in HMC Dockyard Halifax, NS put (thrust) shaft was observed to have by hydraulic jack in four positions of rota- suddenly bent, some two months after a refocused their attention on refitting a tion, each at 90°. By measuring the force steam destroyer at the Ship Repair Unit, minor collision incident. What had hap- needed to lift the shaft against the amount and doing a bit of forward planning re pened? How had this come to be? the shaft actually lifts, it is possible to Christmas leave. A puzzled staff at the Naval Engineer- show any difference in loading in the four The first starshell appeared in early ing Unit recommended Saguenay remove positions. If the shaft is bent, then in the November, and the OPDEF message her port propeller, for the sake of fuel position where the bend is acting down- embodied a definite "Run that by me economy, and return home on one shaft. wards on the bottom bearing, the force again?...." quality. Saguenay reported a required to lift the shaft will be greater. The riddle was set. sudden one-eighth of an inch oscillation of Provided the amount of lift is less than the the port propeller shaft in the engine-room Brainstorming specified bearing clearance, then reaction at the bulkhead gland. Also, oil was leak- testing can be done with the top-half bearing from the gearbox at the gearbox seal, Whilst Saguenay was enroute to Can- ing in place. and the thrust shaft could be seen oscillat- ada the staffs of the Gearing Section at ing at the gearbox output. There was no NDHQ Ottawa and the Marine Systems report of any vibration and all gearing and Engineering Division at NEUA Halifax shaftline temperatures were normal. deliberated independently and then in

MARITIMI- I NG1N1.I RING JOURNAL Dial micrometer arrangement at each of the couplings TURBINE and GEARING

STERN TUBE PROPELLER LOOSE COUPLING

Figure 1. Shaftline General Arrangement

The graphs in Figure 2 show the ery of the compartment towards the centre bracket to enable a clock gauge measure- results. It will be concluded that all is where the plummer block support webs ment of any eccentricity at the very end of reasonably well at the plummer block, but were welded. Was this extra evidence, or the propeller shaft. Once again the video at the thrust block and stern tube there are pure coincidence? Structural platework is camera was used. definite bends. seldom flat anyway! On a very cold December evening a The shaft radial runout (eccentricity) As soon as the Fleet Diving Unit small group of us huddled around a tiny readings arrived next (see Fig 2). (The arrived an underwater video was taken of TV monitor on board the diving tender. readings were taken twice, first with the the hull areas beneath the port plummer As the propeller shaft was turned we shaft turning ahead and then astern.) The block. There was no sign of any structural observed the most startling revelation. data was plotted in two ways — by simul- damage whatsoever. Expectations of this Contrary to all previous experiences of taneously plotting the clock readings at idea being the solution to the problem this kind the instrument showed an enor- 0°, 90°, 180° and 270°, and by plotting evaporated immediately. The underwater mous bend at the tailshaft taper — the position of the shaft in the horizontal camera also showed that there had been no measured at over 0.300 inches. and vertical planes through one full revo- contact between the submarine and the lution. By plotting the locus of the centre ship at all along the port side of the hull. The appearance of the scant damage to the propeller blades had been completely of the shaft, a clear picture emerged as to Significantly, the port shaft and A-bracket deceiving. The impact of the submarine how the propeller shaft was distorted. also appeared untouched —no damaged fin on the propeller, although inflicting paintwork, no marking of the tailshaft It appeared as though the shaft, in the only minor damage to the blades, had fibreglass coating. More perplexed than course of rotation, was moving eccentri- transmitted a massive shock through the before, we turned our attention to the port cally downwards at either end with centre of the screw. The force required to gearbox. respect to the coupling adjacent to the produce such a bend abaft the A-bracket plummer block, which was rotating The main gearwheel teeth looked very was later estimated to be in the order of eccentrically in the upwards mode. At good with no sign of any damage or 800,000 - 1,000,000 Ibf. The relevance first it looked as though the plummer uneven wear. Main gearwheel bearing of the initial dubious poker gauge read- block had possibly been displaced clearances were also checked and found to ings now assumed a leading, if not embar- upwards, thereby causing the shaft to be satisfactory. Strangely, though, the rassing significance. bend, pulling the thrust shaft off centre. thrust shaft only a few feet away from the The Quick Fix Maybe this was the answer. Or were we main gearwheel was bent. The bend of the merely observing the natural droop of a thrust shaft had caused the aft gearbox The ship was docked as scheduled and propeller shaft some distance away from white-metal bearing seal to wipe and another check with a clock gauge con- its main supports? An examination of the score the thrust shaft there. firmed the excessive bend in the tailshaft. deckplates in the plummer block compart- Now we were reasonably confident in our The next day the divers bolted a stiff ment showed them to be uneven. All knowledge as to what was wrong, namely angle-iron extension bar to the port A- tended to bend upwards from the periph- the damage at the tailshaft and the thrust

JANUARY l<)«1 Bearing Reaction @ THRUST BLOCK @ PLUMMER BLOCK @ STERN TUBE BEARING 20 20 19 19

18 18 17 17

16 16

14 15 14 14 " 13 IZ IZ II II io 10 9

8 7 6

5 Indicates shaft 4 Indicates shaft - reasonably straight bent significantly 3 IX 2 I

234367 89 0 2 34567 89 FIGURES IN LBS FIGURES IN LBS FIGURES IN LBS (1000 LBS) 11000 LBS) 11000 LBS) Radial Runout THRUST SHAFT PLUMMER BLOCK LOOSE COUPLING

FORWARD RADIAL RUNOUT

FORWARD INTERMEDIATE COUPLING AFTER INTERMEDIATE COUPLING LOOSE COUPLING (THRUST SHAFT) (PLUMMER BLOCK)

LOCUS OF CENTRE OF PROP SHAFT

-50 -40 -30 -20 -K> -60 -50 -40 -30 -2O THOUSANDTHS OF AN INCH Figure 2. Shaft Distortion Readings

MARITIMF. ENGINEERING JOURNAL shaft. Replacing a tailshaft is a fairly straightforward operation. But the thrust shaft? It would take months to replace all the gearing elements and gearbox bearings. The idea of an in-situ repair immediately jumped the queue of options, and whilst the contractor was in the process of removing the tailshaft we set to work on devising a repair method for in-situ machining of the thrust shaft. If it could be made sure that there were no fatigue cracks present in the thrust Maximum runout shaft, and the main gearing and bearings recorded here had not sustained any damage, then the repair was feasible. The bend started here MAIN GEARWHEEL BEARINGS?3^ We were lucky! X-rays of the shaft were negative and a no-load spin test of the gearbox at 196 rpm revealed no abnor- mal vibrations. The damage was confined to the thrust shaft, affecting both the thrust collar and the output flange (Figure 3). The major snag, of course, was that to machine the thrust collar and output flange a method was required to axially MAIN GEARWHEEL locate the thrust shaft. The fertile mind of our machinery inspector produced an answer which involved two methods, both of which used the main steam turbine to turn the shaft astern. The first method was required to machine the output flange and involved using the bent thrust collar to axially THRUST ** These areas locate the thrust shaft. Although the thrust SHAFT THRUST collar had bent, its thrust bearing was still COLLAR *-* were machined true. Assuming that the high point of the bent thrust collar would track on its true Figure 3. Simplified arrangement showing bearing, axial location could be accom- main gearwheel and thrust shaft plished. The no-load spin test had already with the pads removed. verified there was no axial movement when turning the shaft astern. Once the With most of the brainstorming out of mence. At this stage Saguenay was trans- output flange had been machined the sec- the way we could now proceed with the ferred to the Ship Repair Unit at HMC ond method could begin. nitty-gritty. Checks were carried out on Dockyard Halifax for the remainder of the The second method was required to gearteeth meshing patterns, which proved repairs. This work was done by SRU machine the thrust collar and shaft and satisfactory, and a close look at the plum- staff, assisted by naval personnel from involved connecting the propeller shaft to mer block bearings showed them to be in Fleet Maintenance Group Atlantic. good shape too. Next, the loose coupling the thrust shaft and using the emergency The total indicated runout at the thrust was broken, and the damaged tailshaft trailing thrust bearing in the plummer shaft output flange was in the order of was removed'. An optical alignment check block to axially locate the thrust shaft. A .035 inches. The first job was to machine proved that the A-bracket and stern-tube subtle but important adjunct was, of the output flange face with the thrust shaft bearings were reasonably true and had course, to transfer the thrust pads from the disconnected, then ream the thrust shaft suffered no serious damage. These bear- starboard to the port plummer block. This bolt holes into correct alignment with the ings are huge and can absorb a lot of enabled the pads to tilt in the correct mode intermediate shaft. All this was done with punishment. in order to achieve hydrodynamic lubrica- the thrust pads and cover in place and was tion. Normally the trailing thrust pads A new tailshaft was fitted and coupled, necessary before we could connect the would operate with the shaft turning and the inboard and outboard stern seals output flange. New fitted bolts were ahead, not astern. Also, a 1/2-inch spacer were replaced as a matter of routine. Once manufactured for the coupling, and the was required between the output flange the new propeller was fitted and all the shafts reconnected. and propeller shaft to maintain the thrust fairing plates replaced, the ship was With the gearbox cover and thrust pads shaft position whilst moving the propeller undocked. Now afloat with her fuel out of the way, and lube oil supply to the shaft against the trailing bearing. (It evenly distributed, machining and thrust collar blanked, a portable machine should be noted that the propeller would realignment of the thrust shaft could com- be installed prior to this procedure.) tool was bolted to the gearbox casing.

JANUARY I Acknowledgment The author wishes to express his sin- cere gratitude to Lt.Cdr Neil Latham, Royal Navy, lately of the gearing section at National Defence Headquarters, Ottawa. Steve Dauphinee of DMEE 2, Figure 4. Machining the gearbox thrust shaft in situ. Dr. Jim Matthews of the Dockyard Laboratory, Mr. Doug Nickerson, Figure 4 shows this rig being used to though the thrust shaft had been checked Machinery Inspector at the Naval machine the areas in way of the after oil regularly during the intervening period. Engineering Unit in Halifax, and to those seal and bearing. Turning astern at 50 rpm Whilst several ideas have been sug- staff of the Ship Repair Unit and Fleet under steam, these areas and the thrust gested, none has gained the unanimous Maintenance Group who actually carried collar faces were machined true, then assent of those involved. The most popu- out the repair work to HMCS Saguenay. honed carefully to give a good finish. The lar theory was put forward by Dr. Jim They unreservedly deserve the credit for simplified diagram shows the areas which Matthews of DREA. His premise is that the Saguenay job, and it was a pleasure to were done. With machining operations when the submarine hit Saguenay abaft work with them and to learn from them. complete, the thrust pads were replaced the engine-room the impact actually and shimmed to retain the correct thrust caused the entire ship to bend. The result clearance. The reduced diameter of the was that the bulkhead gland abaft the thrust shaft necessitated fitting an under- gearbox pushed the thrust shaft into plas- size aft bearing seal. All areas were care- tic deformation at its thinnest section fully cleaned out and inspected prior to inside the gearbox. The white-metal bear- flushing the lube oil system. ing seal at the aft end of the gearbox was A successful basin trial ensued and two strong enough to constrain the thrust shaft days later, less than ten weeks from the to run as though it appeared straight. As date she entered drydock, Saguenay was time progressed however, the extra load steaming at full power sixty miles off the on the bearing was sufficient to progres- coast of Nova Scotia. The rest, as they sively wear away the white metal, and say, is history. eventually allow the thrust shaft to run with an eccentricity visible to the naked Non Sequitur eye. This idea, although possessing the Despite the concerted efforts of several highest degree of probability, does not engineers and three scientists, one ques- convince the naval architects who contend tion remains. Why did the thrust shaft that there was insufficient force to cause bend in the first place? the ship to bend at the time of collision. There was simply not enough energy at HMCS Saguenay continues to run the tailshaft during the collision to trans- smoothly nearly two years after the repair mit the required force forward to the gear- was completed. All concerned learned a Lieutenant Woodhouse is the ISL Class box. Moreover, the bend was not great deal from the Saguenay saga, but the Officer at the Naval Engineering Unit in observed until two months later, even mystery still haunts the mind. Halifax.

MARITIMI 1 NCINI I RIM! JOI KNM CANTASS Bringing ASW into the 21st Century

By LCdr Richard Marchand

Introduction System Description The Canadian navy has embarked on to meet Canadian requirements did not The CANTAS system can be divided an ambitious procurement and moderni- exist, so this portion had to be developed. into two major elements: the Wet End, zation program. Part of this moderniza- In December 1984 a contract was let to comprising the array, array receiver and tion process involves the development and Computing Devices of Nepean, Ontario the array handling and stowage equip- production of the Canadian Towed Array for the production of an advanced ment, and the Dry End which comprises Sonar System (CANTASS) which will development model (ADM). This was the processing and display equipment. supply the Canadian navy with a greatly delivered to HMCS Annapolis in Febru- This system division is shown in Figure 1. enhanced ASW capability. This article ary 1988 and has since been undergoing presents a high-level description of the extensive sea trials. The CANTASS Wet End CANTAS system, showing how it will production systems will incorporate The Wet End consists of the following indeed meet the navy's current and future modifications dictated by the lessons four subassemblies: passive-sensor ASW requirements. learned during these trials. a. The Array: The AN/SQR-19 is a Background CANTASS has been designated for neutrally buoyant, 800-foot-long installation in HMCS Annapolis, HMCS array measuring 8 centimetres in Experimental towed arrays have seen Nipigon and the twelve Canadian patrol service in the Canadian navy since the diameter. The neutral buoyancy is frigates. mid-1970s. The success of these experi- necessary to ensure that it remains as horizontal as possible while mental systems motivated the initiation of Requirement being towed. The array consists a full-scale development project to of a number of modules which are produce a tactical towed array sonar sys- The stated requirement for CANTASS coupled together in the configura- tem for use on Canadian warships. Treas- is to provide a passive means of detecting tion detailed in Figure 2. A gen- ury Board approved the development and tracking a hostile surface or subsur- project in September 1983, and in April face target at a distance that exceeds the eral description of each module 1984 gave preliminary approval for the target's anti-ship-weapon release range. follows: production of Canadian systems. The requirement was highlighted in the (1) The Drogue at the aft end of the array performs a similar Early in the project it was decided that white paper on defence (1987) which specified the need for continued surveil- function to that of a sea Canada would buy the most up-to-date anchor. The drogue damps array sensor available, the USN's lance and protection of Canadian, North American and North Atlantic Treaty out any whipping action of AN/SQR-19. Purchasing off-the-shelf the array as it is being towed equipment would minimize the technical areas. Canada's capability for the surveil- lance of these ocean areas will be greatly and provides the necessary risk and overall development time frame. drag to ensure the array A suitable display and processing system enhanced by the use of CANTASS. remains horizontal during operations. The drogue consists of 75 feet of 3/4-inch polypropylene rope. (2) The Telemetry Drive Mod- ule (TDM) is a multiplexing RECEIVER PROCESSING and DISPLAY module that digitizes and then time series multiplexes the acoustic data produced by the various acoustic I I modules and the non- PROCURED FROM USN CANADIAN DEVELOPMENT acoustic data produced by the HTDM. Figure 1. CANTASS Wet End and Dry End Division (3) The Heading, Temperature and Depth Module

JANUARY IW (HTDM) performs two Frequency, Low Frequency receiver manufactured by Gould functions. It measures the and Medium Frequency. Inc. of Baltimore, Maryland. To array's magnetic heading, Although the AN/SQR-19 minimize schedule and technical sea temperature and array array has two high- risk it was decided to have Gould depth, and performs a mul- frequency modules, in Inc. produce the array receiver tiplexing function for the CANTASS these are used because of their expertise in VLF modules of the array. in the medium-frequency producing the arrays. All this data in turn is for- mode. The hydrophone The array receiver accepts the warded to the TDM for spacing in the modules array acoustic data and the non- onward transmission to the determines the highest acoustic data (temperature, array array receiver. operational frequency of the heading and array depth) and puts (4) The Vibration Isolation array (to maintain the direc- it in a usable format for the CAN- Modules (VIMS) act as tionality of the array TASS shipborne electronic sys- shock absorbers that damp beams), and the array's tem (SES). The array receiver out any tendency of the acoustic aperture basically also provides the array with a array to snake through the sets the lower frequency constant D.C. power supply. limit. Figure 2 shows the water. They also provide d. Data Recorder: The data vibration isolation of the acoustic aperture of the AN/SQR-19 array. recorder is a high-density digital array from the tow cable. recorder (HDDR) unit used to This isolation reduces the b. W/nc/i: The array winch assem- collect the raw acoustic data col- tow-cable induced noise bly consists of the USN's OK-410 lected by the array. The HDDR level that the array would winch system (shown at Figure 3) does this by recording the acous- see. The four modules which has capacity for holding tic information provided to the described thus far make up 5600 feet of tow cable plus the front end of the array receiver. the non-acoustic section of array. Since the cable is nega- This provides the ship and shore the array. tively buoyant, array depth can be establishments the opportunity to (5) The acoustic modules of the controlled by varying the amount play back data for analysis and array consist of equally of cable streamed and the ship's training. spaced hydrophones to speed. Dry End cover the frequency bands c. Array Receiver (AR): The array of concern; the bands of receiver is a Canadian-developed The Dry End, or shipborne electronic interest being Very Low system, consists of the six subassemblies shown in Figure 4.

VIM VIM TDM LFM LFM urn HfM HTM MFM LF~M LFM VLF VLF VLF VLF VLF VLT vi_r VLF HTDM 11 «a • 1 *s • i •1 •2 «2 •3 ft 4 •I •2 H3 N4 »5 •6 «7 •e

ARRAY ACOUSTIC APETURE

Figure 2. AN/SQR-19 Array Functional Block Diagram

10 MARITIME ENGINEERING JOURNAL a. The Data Management and Dis- tribution Unit (DMDU): The OPERATOR PROTECTIVE SCREEN DMDU is a multiprocessor-based GROUND CONNECTIONS unit that contains 32 Mbytes of \D ASSEMBLY mass memory. The DMDU contains two display processors, a system controller and a tracking processor, all of which are based on Digital Equipment Corpora- tion's Micro J-ll processor design. All four processors operate on a ring bus configuration and all are capable of accessing working memory through the system controller. The function of the DMDU is to provide the SLIP RING

switching functions necessary to WINCH transfer data between the signal

processing unit and the displays HANDLING DRUM via the DMDU's working memory. Figure 5 shows the DMDU MAINTENANCE KIT system architecture. A brief functional description of each proces- Figure 3. Typical Handling and Storage Gear Compartment Location sor follows: (1) The System Controller's main purpose is to control the sequencing of operations in the SES. The principal operations are: — The orderly sequencing V. of start-up and running HCU of the system. SYSTEM SWITCH POWER c"-^ — The gathering and distri- CONTROL UNIT HARD COPY bution of data within the UNIT c SES. — The monitoring of system performance and the reporting and locating of system faults. — The calculation of a fig- SYSTEM DISPLAY ure of merit (FOM) for POWER DRIVE CONTROL the CANTAS system. ASSEMBLIES UNIT DISPLAY DATA ARRAY DRIVE MANAGEMENT SIGNAL TAPE This FOM will give the PROCESSOR RECEIVER RECORDER ASSEMBLIES DISTRIBUTION UNIT MAINTENANCE operators a range perfor- CONSOLE mance estimation for the BACK-UP POWER MC system. SWITCH — The estimation of a fine bearing of incoming target signals of interest. This fine bearing algorithm will give a more accurate bearing of the target. Without this CPF UCCS/265 ADLIPS —I process the target's SHIP MOTION direction would only be accurate to plus or minus Figure 4. CANTASS Shipborne Electronic System several degrees. Functional Block Diagram (2) The Display Processors (DP) have several functions to perform. The principal functions are: — The management of operator inputs.

JANUARY IIR'I 11 — The display of environ- facility can be overridden or involved in the processing. The mental, alert and status complemented by operator data is first divided into overlap- information. entry of targets of interest. ping blocks and transferred into — The management of the frequency domain through the tabular area displays. b. The Signal Processing Unit use of the Fast Fourier Transform — The management of the (SPU): The final production units (FFT). Next the data is beam- acoustic area displays. of CANTASS will make use of formed, hetrodyned, decimated — The management of the the militarized AN/UYS-501 sig- and brought back to the time SHINPADS display nal processor, although the CAN- domain using the Inverse Fast input and output. TASS ADM makes use of a Fourier Transform (IFFT). Suc- commercial version of this signal — The performance of cessive blocks of data are con- processor built by Motorola. The start-up, diagnostics and catenated so that the data again equipment health AN/UYS-501 is an eight- represents a continuous time monitoring. arithmetic-unit (AU) processor series signal. The data is again utilizing four megawords of transformed using the FFT. The (3) The Tracking Processor's memory. It has a throughput of sole purpose is to determine intermediate step of returning to 320 million floating-point opera- the time domain allows for if any discrete frequency tions per second. components on any given increased frequency resolution. The role of the signal processor is bearing exceed the detec- Finally the data is integrated, tion threshold, and if so to to process the time domain acous- scaled and placed in the working initiate automatic tracking. tic data into frequency domain memory of the DMDU for dis- The automatic tracking data ready for presentation to the play and track processing. The operator. Several stages are

LOWER UPPER UPPER SCREEN REEN LOWER SCREEN NTDS FAST HSSI

DISPLAY DISPLAY PROCESSOR 1 PROCESSOR 2

ADLIPS

MAINTENANCE ft' CONSOLE UYS-501 SIGNAL PROCESSOR Figure 5. DMDU System Architecture

12 MARITIME i NGINI:I:RING JOURNAL signal processing data is illus- not provide the required resolu- capability to maintain hardcopy trated in Figure 6. tion. The operator can switch archival information on targets of c. The Display Driver Assembly between monitors for display and interest. tracking purposes. CANTASS (DDA): The DDA provides the System Capabilities interface between the DP in the makes use of two DSDs. CANTASS is a passive system which DMDU and the display monitors. e. The Maintenance Console (MC): is capable of detecting current surface and They provide for the display of The CANTASS ADM MC is an the appropriate graphics and AN/UGC-504 teleprinter unit. It subsurface targets out to the second con- vergence zone and beyond. CANTASS acoustic data on the SHINPADS connects to the DMDU and allows will therefore enhance a ship's ASW displays. The alphanumeric and the maintenance technician to capabilities by filling the void in the detec- graphics data is received by the observe equipment health DDA over an NTDS fast parallel monitoring reports, or to enter the tion envelope as shown in Figure 7. CAN- interface. The acoustic video diagnostic mode and perform TASS takes the raw acoustic data and forms 43 equally spaced beams in arc-sine information is received on its fault isolation tests. The MC can space. Several of these beams are illus- high-speed serial interface also be used to set the system trated in Figure 8. The beamformed (HSSI). The DDA also accepts adaptation parameters in the sys- acoustic data is presented on the two dual- the operator input from the con- tem's EEPROMs; these screen SHINPADS standard displays, trol unit display (CUD) and for- parameters establish the default from which the operators can determine mats it for use by the DMDU's values that the system uses when display processors. it is first powered up. The CAN- discrete narrowband frequencies and TASS production systems will broadband signatures associated with a d. The Dual-Screen SHINPADS target, as well as determine a target's Standard Displays (DSDs): The make use of a microcomputer as the MC. bearing. Through the use of a fine bearing DSDs are modified SHINPADS algorithm, an accurate bearing can be standard displays. To meet CAN- f. The Hardcopy Unit and Video passed to the command and control sys- TASS display requirements, two Switch: The SES has a video tem (CCS). Target-motion analysis can monochrome CRTs (rather than switch which allows for the con- then be carried out by the CCS to deter- the standard single-colour CRT) nection of one of four DDA video mine a range to the target. are housed in the DSDs. The outputs to the Honeywell monochrome monitor is used VGR-5000 hardcopy unit. The The system and/or operator can create because of the requirement for VGR-5000 produces an 8 1/2" x and track up to 240 tracks and markers high resolution for displaying 11" black and white hardcopy and assign these to a maximum of 99 tar- LOFARgram data. The current printout of the selected screen. gets. The system will automatically track colour monitor technology does Thus the operator will have the

• LOG SCALE. 4 BIT PACK MAGNITUDE NOISE & 1, • LINEAR SCALE. 8 BIT PACK NORM INTEGRATE » 4:1 OR LOC SCALE. 4 BIT PACK

LOG SCALE. 4 BIT PACK

LINEAR SCALE. 8 BIT PACK

4.1 OR LOC SCALE. 4 BIT PACK

LOG SCALE. 4 BIT PACK

LINEAR SCALE. 8 BIT PACK

4.1 OR LOC SCALE. 4 BIT PACK

Figure 6. Signal Processing Data Flow

13 these targets until they become faded, lost or are deleted by the operator. The system also possesses an equipment health monitoring function that tracks and reports the status of all system subcomponents. When faults are reported, the system maintainer can make use of a built-in diagnostic program to isolate the failure down to the lowest repaira- ble unit (LRU) — in most cases, a circuit board. The software design methodology for CANTASS was based on existing software design concepts within DND. The modular construction of the software tasks, combined with a global coefficient data base for use by all processors, makes the system relatively easy to update with new processing algorithms. This will allow the system to be updated to meet the ever-changing threat. Figure 7. CANTASS Areas of Coverage

NOTE: ACTUAL TOWED ARRAY BEAMWIDTH DEPENDS ON FREQUENCY.

FORWARD AFT ENDFIRE ENDFIRE

LCdr Marchand is the CANTASS project engineer at NDHQ. AFT CONICAL BEAM FORWARD CONICAL BEAM BROADSIDE

Figure 8. AIM/SQR-19 Beam Shape

WRITER'S GUIDE We are interested in receiving unclassified submissions, in English or French, on subjects that meet any of the stated objectives. Final selection of articles for publication is made by the Journal's editorial committee. Article submissions must be typed, double spaced, on 8!/:" x 11" paper and should as a rule not exceed 4,000 words (about 17 pages). The first page must include the author's name, address and telephone number. Photographs or illustrations accompanying the manuscript must have complete captions. We prefer to run author photographs alongside articles, but this is not a must. In any event, a short biographical note on the author should be included with the manuscript. Letters of any length are always welcome, but only signed correspondence will be considered for publication.

14 MARITIME ENGINEERING JOURNAL Development of a Reverse Osmosis Desalination System for the Naval Environment

By Morris Shak and Real Thibault

Introduction RODS vs Evaporators Reverse osmosis desalination systems Past policy dictated that the navy use ent concentrations are separated by a (RODS) are not a new concept. They have evaporators to provide potable water and semipermeable membrane, then the been utilized and maintained in land and boiler feedwater. As RODS improved and purer, less concentrated fluid will pass marine installations for more than two positive claims were made by other through to the more concentrated side. decades. Over the years these systems marine users, their appeal increased, This action will continue until concentra- have been improved continuously through especially in view of the fact that operations equalize or the pressure on the more the development of better semipermeable tional costs were expected to be 50 per- concentrated side becomes high enough to membranes that can withstand higher cent lower than for conventional prevent any further flow. The minimum pressures, yet research and major design evaporators.'-2 pressure that prevents further flow of the enhancements continue in search of an solvent is called osmotic pressure of the Reference 3 outlines in particular the improved product. solution. RODS application for the Tribal class. It In May 1984 the Department of Supply states, "The development of Reverse and Services contracted a Canadian Osmosis Desalination (ROD) has changed With the development of suitable syn- manufacturer to build a reverse osmosis the design authority perspective on thetic membranes to withstand the high desalination system, which would incor- domestic steam systems, and has also pressures, and enhancements to the porate a high-pressure energy recovery generated discussions concerning the mechanical attachments to hold the membranes in position under pressure, revers- pump, for testing and evaluation on a replacement of steam with electric heating Tribal-class ship. The scope of the techni- throughout the ship. A SHIPALT package ing the flow by reverse osmosis could be realized. This procedure now permits cal requirements essentially attempted to is being staffed to revise the existing desalinating highly concentrated sea keep the design simple and modular with DDH-280 evaporators and replace them materials and components compatible for with RODs. Ships using reverse osmosis water down to the saline concentrations acceptable for potable water and boiler the naval environment. will have a significantly reduced steam demand and will be able to operate on one feedwater. Early membranes, originally In June 1984 DMEE 4 tasked the Naval auxiliary boiler instead of two in all but made from cellulose acetate, have been Engineering Test Establishment (NETE) superseded by polymer esters which pro- the most extreme conditions. In this con- to outline a test program that would: vide for higher reject rates and are less figuration, availability of domestic steam susceptible to biological fouling. a. prove the equipment integrity can be assured by two auxiliary boilers Presently, available membranes are found (shock test); without a third source of steam being required." in two basic designs, hollow fibre and b. verify equipment structureborne spiral wound. and airborne noise; Finally, the emphasis on the impor- Selecting a ROD System c. set up SOAP on the HP energy tance of reverse osmosis in warships is recovery pump; and documented in various articles.4-5 The ROD systems in general are packaged d. confirm performance by a overall reduction in energy and main- in various configurations and sizes to 200-hour endurance run. tenance costs in addition to the space meet specific installation demands and a saved, safer ambient temperature, non- wide range of throughput capacities. The The complete RODS was delivered to polluting operation and the more con- system design selected in 1984, and as NETE for evaluation in late 1984. Had it venient RODS package have been the contractually configured for the Tribal- satisfactorily met all the test program prime reasons for the decision to phase class evaluation, is outlined in the flow requirements, it would have left NETE in out evaporators in favour of RODS. diagram depicted in Figure 1. spring 1985. However, this was not the case. Deficiencies in operational Reverse Osmosis In this arrangement, sea water is fed to a filter media tank with a boost pump. The parameters and structural packaging To understand reverse osmosis, a basic filter media tank acts as a coarse filter necessitated a series of modifications to be appreciation of osmosis is required. removing most of the undissolved or sus- undertaken. The redesign, repackaging Osmosis is the process of diffusion pended material in sea water. An electric and retesting are still continuing and this through a semipermeable membrane. It article attempts to trace the developments dictates that when two solutions of differ- heater prevents the water from freezing to date. and increases the efficiency of the RODS at temperatures below 5°C. A follow-up

JAM \KY ]')«'> 15 LEGEND

H PRESSURE SWITCH cJ] REGULATWG WLVE (V)TOTALIZIN GMETE R

P1 VACUUM BREAKER 6* CHECK WLVE 0 TEMPERATURE SWITCH

§3 SOUNOO VALVE J RELEF VALVE CM GATE VALVE

FUER BACKWASH TO WASTE (P) 60-70 US GPM

2U US GPM WASTE J J ,— MEMBRANE CLEANNO CONNECTIONS \ ', 275 US GPM '-a \ 31.300 PPM TDS-^ 4 SE E NOTE ^

1 ' po . / POTABLE WATER \C FIRST STAGE \ M HfiH PRF<«1JRF \ 8800 US GPO 1 ENERGY RECOVERY ~A 1 fcs=j IX1~ / 11 rB3 BLEEDV '" \ \U 1 ® S 0 \ 0*P 1 / X<3 (==- 00— /

5E MEMBRANES -KW— • '—* — iSh-(^ ) MEDIA \ /CARTRCGE \^_Js u | • DO- X * POTABLE i P) 1 WATER AGE PRODUCT DRAI' N i( -1@ n NE/LVW . ' ii Cb- gj™GPO n 1= ^HW-1 ! _} lISo* « TDS HYPOCLORINATOR TANK Y ^^-^ i FEEDWATER .AND FEED PUMP

^-.H^UJ IWBI - 1* .- SECOND STAGE MAKE-UP ( 1 ) BOOST PUMP ,/ L_——-^" ^ 10 TONNE/DAY ^--^ REJECT TO FRST STAGE —- ^ 26«> US GPO 5 1 TUNNE/UAT . ~ 350 PPM TDS 880 US GPO (°) * /a <> » ^RECYCLE 1 >^ i.3 TONNE/DAY „ 1 ric, [ 1.560 US GPD *°TL ~ loso PPM TDS

IW UIKtb/MNUIt ~5/l) PtTl lUb V 7 T" SECOND S AGE MEMBRANES pj 0=^-—-Q C|l -O« Q()

0.50 J FITER BACKWASH) SECOND STAGE i ' !c i 1 BOILER HIGH PRESSURE PUMP TV FEED / tL WATER ^jj. \-MEMBRANE CLEANING CONNECTONS -> SELOND STAGE PRODUCT 67 TONNE /DAY Y= RECOVERY RATIO FIRST STAGE fi FIXED 1760 US GPO SECOND STAGE CAN BE VAR(D FROM 025-075 BYALTERMJ RECYCLE Figure ] . Flow diagram for the L>DH-280 ROD system ~50 PPMTDS

5-micron cartridge filter further removes semipermeable membranes. The highly obtained when supplied with sea the finer particles that pass on through the concentrated brine is rejected while the water at 35,000 mg/L total dis- filter media tank. potable water, with an acceptable saline solved solids at 25°C; concentration, now either flows entirely c. the plant shall be operational with The accumulator before the high- into a holding tank, or part of it is used pressure (HP) pump reduces the pulsa- sea water at -2.2°C to 31°C; as intake to a second-stage system using tions caused by the HP pump suction, as only two spiral wound membranes in one d. the post-water-treatment system well as the higher frequency pulsations housing to provide boiler-quality feed- shall ensure that potable water caused by the boost pump. An HP pump water with much reduced salinity concen- produced is suitable for human is used to deliver sea water through tration. consumption and that it meets the another accumulator to the first-stage standards specified in A- semipermeable membranes at 800 psig. MD-213-001/FP-001, QSTAG Performance The relief valve intended to protect the HP 245 and STANAG 2136; pump and the semipermeable membranes In the envisaged application, the per- is set to 1000 psig. e. the second stage shall be capable formance of the ROD system was defined of producing 20 percent of the Twelve 4"-dia. x 40"-long spiral in the Technical Statement of Require- first-stage capacity specified; and ments as follows: wound semipermeable membranes con- f. the first-stage effluent for drink- tained within the membrane housing are a. the rated capacity of the plant ing water must contain less than configured in two parallel flow branches. shall be 33.3 cubic metres (8,800 500 ppm chloride, while the Each branch is made up of three mem- U.S. gals.) of fresh, potable water second-stage effluent for boiler brane housings in series with each mem- daily; feedwater must contain less than brane assembly containing two b. the rated capacity shall be 7.3 ppm chloride.

16 MAR1TIMI-: FNGINKKRING JOURNAL Design water and boiler feedwater with stainless steel should be used. Galvanic minimal valving. corrosion must be reduced in general and, To effectively design a ROD system in particular, metallic particles must be for use on board a naval vessel, manufac- It should be noted that producing boiler kept away from the membranes. Rein- turers must take into account the unique feedwater requires a two-pass system to forced plastic and rubber hoses do not operational and environmental require- meet water qualities. One way of accom- generally cause any difficulty in applica- ments of all applicable military specifica- plishing this is to direct the product of the tion. But, impregnated fibreglass tions. The following were the major first pass to a holding tank, isolate the sea- filament-wound vessels must be used with design guidelines imposed on the con- water supply, then draw the potable water caution by first verifying their structural tractor : from the holding tank for a second pass integrity. Although all the hardware — Modular design to allow the unit to through the ROD unit. The alternative required to put together a commercial be retrofitted into an existing ship method, followed by DND, is to provide RODS may be available, it is extremely and to enable subsequent modifica- a second-stage RO unit within the main difficult to find materials suitable for use tions to be accomplished without ROD system. Apart from being a simpler, in the naval environment. Even then, redesigning the complete unit. faster and cheaper operation, this method delivery may take weeks or months. allows the reject output of the second- - Allow for easy access and main- stage pass to be reintroduced to the sea- The First RODS Prototype tenance. water inlet to reduce the saline content of The first RODS prototype which - Restrict size to accommodate the the seawater supply, thereby improving arrived at NETE in late 1984 was com- geometry of the currently fitted the efficiency of subsequent first passes. posed of four skids as illustrated in Figure evaporators, including maintenance 2. It included: space. Finally, the contractor was instructed to provide items such as drip trays to cap- a. a Control Skid with a total weight — Keep weight to a minimum. ture and divert the condensation forming of 2165 Ib, comprising electrical — Satisfy applicable shock, noise, on the RODS, suitable operational cabinets, pressure gauges, salin- vibration and inclination specifi- monitoring and product water quality dis- ity monitors, second-stage pump, cations. play instrumentation, and to use accepta- motor and accumulator, 5-micron ble materials. Of primary concern were — Include an energy recovery pump filter, and solenoid valves on one materials that would withstand the rigors portion, and heater, product flow- for the high-pressure first-stage of a shipboard environment both inter- pump (for a saving of 50 percent of meters and suction pump and nally and externally. energy costs). motor on the other portion; - Allow a change-over from a one- Seawater handling does pose problems b. a Filter Media Tank weighing stage to a two-stage operation, and in the selection of materials. Copper- 3260 Ib with its four ball valves; vice versa, with minimal valving. nickel, nickel-aluminum-bronze and c. a Membrane Skid, complete with high-chromium/high-molybdenum stain- — Produce both domestic potable first-stage accumulators, having a less steel are desirable, especially in areas total weight of 1800 Ib; and of little or no flow.7-8-9 Fasteners of 316 d. a First-Stage Skid on four shock mounts, with associated pump, motor and V-belt drive, weighing 1310 Ib. The four skids, when assembled together, were isolated on eight 1000-lb vibration mounts. The complete assembly measured 66" wide x99" long x 82 3/4" high, with a total weight of 8535 Ib. Prior to arriving at NETE the RODS had passed various functional tests at the manufacturer's facilities (supervised by DMEE4). However, the prototype failed the structureborne and airborne noise tests which were conducted at NETE in January 1985. Although the airborne test results were close to the allowable limits, they did not meet the specification in all the frequency bands. The structureborne test data demonstrated that the system was especially inadequate in the 32-Hz and 64-Hz frequency bands. As a result of these tests, the RODS supplier hired BBN Laboratories Inc. of Cambridge, MA to prepare a proposal for improving the structureborne noise performance. In Figure 2. The first RODS prototype May 1985 BBN Laboratories issued a

JAM'ARY I1K9 17 technical memorandum to the supplier, Cradle d. Accumulators who in turn requested NETE through Remove the existing shock Change the high-pressure DMEE 4 to modify the RODS mounts and bolt the RODS to a accumulators to oversized accordingly. supporting cradle. Mount the new accumulators for better vibration RODS assembly on four 6E-2000 attenuation. Since the performance of the original shock mounts located about the RODS during the structureborne vibra- RODS' centre of gravity. (Since By December 1986, the reconfigured tion test was extremely poor, it was the original recommendation RODS had passed both the airborne and important to gain a better perspective of called for placing the mounts 80" structureborne tests. Following the suc- the final expected structureborne results apart, while the RODS is only cessful tests, the RODS was disassembled of a shipboard installation before 66" wide, the final configuration into its four major skids and cradle, and implementing changes. The three factors was modified. The RODS is still the physical changes were measured and of concern in structureborne noise deter- supported at four locations about recorded. Two of the reworked skids mination were the equipment, its mount- the centre of gravity, but two from RODS II are illustrated in Figure 3. ing arrangement and the impedance of the 6E-1000 mounts are used at each ship's interface structure. In August 1985 location.) the impedance at the proposed RODS mounting location on board HMCS Athabaskan was measured. With favoura- ble impedance test results, all parties concerned were optimistic that incorporating the proposed modifications would permit the new system to satisfy the structureborne criteria. The Second RODS Prototype The conversion process from RODS I to RODS II continued up until November 1986. Incomplete drawings, unexpected interferences in the structural steel members, on-the-spot redesigns and protracted part deliveries played major roles in the long delays. The modifications recommended by BBN Laboratories included the following: a. First Stage (1) Add the accumulators to the first-stage skid since the skids and the motor/pump assembly had high vibration levels. (2) Use a softer shock mount. (3) Place all four mounts above the column members of the membrane skid. (4) Move the horizontal centre of gravity of the first-stage skid to the horizontal centre of the shock mounts in order to minimize roll response. b. Second-Stage Pump Place the second-stage pump/motor assembly on the vibration mounts. (The original design called for three GE-100 mounts at the base and a 7M50 sway mount acting through the centre of gravity and providing restraint in the athwartship direction. This design was subse- quently changed by incorporating two base mounts and two sway mounts.) Figure 3. The first-stage and membranes skids of the second RODS prototype on the medium-weight shock test machine.

18 MARITIME ENGINF.I-RING JOURNAL By this time many drawings were quite The forced-vibration test started in a. The filter media tank skid was different from the original design draw- August 1987. By September, the lack of structurally reinforced by adding ings because of errors, general improve- adequate structural rigidity to meet mili- angles to the two unbraced sides ments, and modifications to satisfy the tary forced- vibration specifications led to of the skids. These angles were structureborne noise specifications. Also, the failure of the control skid and the filter designed to be easily removed for the original design of the RODS did not media tank skid. The other two skids maintenance. lend itself to the potential abuse that hard- encountered similar problems. Conse- b. The membrane skid, which is the ware is expected to withstand in a naval quently, the design and fabrication of a most rigid of the three, incorpo- environment. Installation of protection third prototype incorporating extensive rated heavier brazing of pipe shelves and bars, and the rerouting of structural stiffening was undertaken. joints and additional structural delicate items had to be undertaken, and The Third RODS Prototype reinforcement to prevent the some equipment had to be reoriented to membrane pressure vessel vibra- better suit the RODS location with respect The major improvements incorporated tion pads from sliding out. to ship space. on the skids in the RODS III version can c. The first-stage skid frame was be summarized as follows: found to be structurally weak. The high-pressure pump, complete with energy recovery pump, resonated with an amplification factor greater than ten. The motor resonated with an amplification factor greater than five. The modifications consisted of changing the angles to heavier channels under the pump and placing gus- sets in the proper locations. These changes reduced the skid vibrations to a tolerable level. d. The most complex situation developed at the control skid (Figure 4). Some of the problems encountered and attempted reme- dies were: (1) The heater thermostat did not meet military standards. It failed and was replaced with a military-standard unit. (2) The heater elements vibrated quite noisily at various frequencies until a res- trainer was introduced. (3) The four spot-welded bolts holding down the electrical control cabinet were severed from the cabinet. Special arc welding was used for the four bolts while a sway support was used for the panel inside the electrical control cabinet. (4) The needles (pointers) of the various gauges and monitors were difficult to read due to excessive oscil- lations. Structural steel was added to the frame for extra stiffness. There was a substantial improvement in the behaviour of the first three skids when subjected to vibration tests. How- Figure 4. The control skid of the third RODS prototype on the MB-C60 shaker.

JANUARY l'W> 19 ever, although the control skid had also gravity of each of the two electrical cabi- model 1122D pump. Using this standard improved, its performance under forced nets was significantly lowered. The light pump and a 25-h.p. electric motor in lieu vibration remained unacceptable and the gauges were relocated above the electrical of a 10-h.p. motor, the ROD system could heavy bracings and gusseting left much to cabinets while the massive salinity moni- produce 12 320 U.S. gpm versus 8 800 be desired in terms of appearance and tors were moved below the pressure U.S. gpm. Since ROD units are tempera- accessibility. gauges. Finally, the frame was fabricated ture dependent, this increase would allow of heavier channels and reinforced along one unit to supply the water requirements The Fourth RODS Prototype the weak axis to improve structural resis- even in northern waters. To achieve It must be noted, at this point, that tim- tance for the second-stage pump, heater RODS V, minor changes had also to be ing was becoming a critical factor. As the and filter housings along both athwartship made to the plumbing to convert to a RODS designed for the DDH-280 class directions. The results in appearance, three-parallel fluid circuit, and a throttle also met the requirements for SRP II, the maintenance accessibility and vibration valve had to be installed to maintain the second batch of Canadian patrol frigates, and shock resistance were dramatic. The required operating pressure. total new wet weight for the RODS IV it was decided to have the unit tested and Conclusions approved by December in order to meet increased to 9250 Ib, plus 1380 Ib for the the SRP II contract schedule. By October cradle. It can now be seen that the long road taken to design this particular RODS was 1987, shock testing remained outstand- The modified RODS IV (Figure 5) principally due to the manufacturer's ing. The modified drawings had to be fur- passed the shock tests in November 1987. unfamiliarity with military specifications. ther revised to reflect the latest level of This was not, however, the end of the Unnecessarily high centres of gravity, fabrication and this test alone was esti- design. The four modular skids still had unbraced light structural steel construc- mated to take approximately three to be fitted on the cradle. The unit eventu- tion, the use of non-approved military months. ally passed all the required tests, and the components, inferior vibration mounting drawings were provided to PMO SRP II The basic engineering principles of the system and hard-mounting (second stage) in time for inclusion in the project. components arranged on the control skid when vibration mounts were required had been ignored in the original design. The Fifth RODS Prototype were some of the major errors committed The top frame was of light construction, in the original design. Lack of apprecia- with little gusseting or bracing. The two Further modifications have been made tion of the problems inherent in shipboard heavy electrical cabinets were located too to produce a RODS V configuration, operation was evident in the original high, while in one electrical cabinet the enhancing the performance of the unit design. heaviest item was placed at the top. with minimal changes to the unit as originally designed. An increase of approxi- The final package is the first modular At the risk of missing the December mately 40 percent in freshwater RODS that qualifies for use in the naval deadline, a decision was made to com- production is feasible if the energy recov- environment, making it extremely con- pletely redesign and rebuild the control ery pump is not utilized. The energy venient for installation, refit and modifi- skid. Remarkably, the work was com- recovery pump was based on an FMC cations. It is also the first such system that pleted in only three weeks. The centre of can provide boiler feedwater in one simple, continuous operation in any proportion to potable water. Finally, it is also the first such system which employs an energy recovery pump. On certain ships, this feature will prove extremely useful. The experience with RODS has clearly demonstrated that with increased cooper- ation between manufacturers and test authorities at the early design stages, suitable and efficient products can be developed for naval use. The need to test and evaluate equipment before installation on naval ships has been shown to be an integral component of the procurement and SHIPALT process. Acknowledgments The authors gratefully acknowledge the support of those members of the NETE organization who actively partici- pated in the many modifications and trials associated with the RODS project. In particular they would like to acknowledge the dedicated support of Messrs J. Ouellet, J.C. Demers and J. Jansma. During two and a half weeks in late 1987 these gentle- Figure 5. The fourth RODS prototype

20 MAR1TIM1 I NCINI I RING JOURNAL men drastically modified the RODS con- 7. Tuthill, A.H./Schillmoller, C.M.: Guide- 9. Hornburg, C.D./Morin, D.J.: Design trol skid without the aid of drawings. lines for Selection of Marine Materials; Studies and Requirements for 5.0 MSF and Finally, thanks are owed to Mrs C. Hen- New York, N.Y., Inco (May 1971), 2nd RO Type Desalination Plants; Vol. I. derson and Mrs. D. Laberge for their Edition, PP. 1-38. Technical Proceedings, WSIA 10th Annual patience whenever the typescript came to 8. Aluminum Bronze Alloys Corrosion Resis- Conference and Trade Fair, Topsfield, them with yet another change. tance Guide; The Aluminum Bronze Advi- Mass: Water Supply and Improvement sory Service. AB Publication No. 80. Association (July, 1982). Herts, England: Copper Development Association (July 1981) PP. 1-27.

References 1. Sakinger, C.T.: Energy Advantages of Reverse Osmosis in Seawater Desalina- tion; Desalination. 40 (1982) 271-281 — Elsevier Scientific Publishing Company, Amsterdam. Printed in the Netherlands. 2. Manuel Morris: A Comparison of Desalt- ing Process from an Energy Point of View; Desalination, 40 (1982) 237-244 — (See Reference 1). 3. Scholey, J.R./Neri, J.P./Mueller, G.W.: DDH-280 Waste-Heat Recovery System; Maritime Engineering Journal, Winter 1985. 4. Sessions, M.P.N.: The Efficient Use of Energy in Warship Auxiliary Systems; Journal of Naval Engineering, Vol. 29, No. 3. June 1986, PP 508-518. Morris Shak is a senior project engineer at the Naval Engineering Test Establish- Real Thibault is a technical officer with 5. Blakey, R.G. : Auxiliary Energy What ment in Lasalle, P.Q. DMEE 5. From May 1986 until February Price to Pa\l; Journal of Engineering, 1988 he was the NDHQ project officer for Vol. 30, No". 1 PP. 193-201. RODS. 6. Gooding, Charles H.: Apply the Membrane Advantages; Chemtech, June 1985.

Advanced Marine Engineering Course Projects A Submarine Propulsion Gas Turbine Design Tool Transient Performance Health Monitoring by Lieutenant Commander R. H. Gair, CF by Lieutenant Commander N. T. Leak, CF

Summary Abstract The design of a total system has always This report investigates the feasibility been one of great complexity. This paper of gas turbine transient performance anal- analyzes design theory, looks at modern ysis as an engine health monitoring techniques of applying the theory and then (EHM) technique for a typical simple uses this information to solve the subma- cycle, twin spool marine gas turbine main rine propulsion problem. The solution propulsion engine. A computer modelling uses an expert system as a decision- program was used to simulate transient making tool. The final product is a com- performance under fault and no-fault con- puter program which provides the ditions. The report concludes that tran- designer with a comparative analysis of sient performance analysis exhibits all solutions for all design considerations. sufficient potential as an EHM technique to warrant further development, particu- (Royal Naval Engineering College, larly as a complementary performance Manadon, July 1988) analysis technique within an integrated EHM system. (Royal Naval Engineering College, Manadon, June 1988)

JANUARY I98") 21 The "Black Art" of Propulsion System Alignment

By LCdr B.B. Staples, PhD, P. Eng

Introduction The process of propulsion system alignment in a warship has been an area of mystery and confusion to most naval officers (even, marine systems (MS) officers) due, basically, to a lack of exposure or experience. Certainly, during the working careers of most MS officers, the opportunity to "oversee" a complete installation is limited. The purpose of this paper is to help eliminate this confusion by describing what constitutes the process called alignment, some of the alignment techniques available to the shipbuilder, and what is being done in the Canadian Patrol Frigate (CPF). Background The process of propulsion system nology "shaft alignment" is a misnomer. ingress of sea water. It must also be suffi- alignment includes the alignment of the Since the shaft is supported by bearings, ciently flexible to accommodate the "hog propeller shaft system to the ship's hull, the process is, in fact, one of bearing and sag" movement of the ship's hull. followed by the gearing to the propeller alignment. By moving the bearings either shaft system and finally the engines to the forward or aft, vertically or athwartships, The correct alignment of the propeller gearing. This article will speak to the the alignment of the propulsion shafting shaft is essential in preventing the following conditions: more intricate of these alignment tech- system can be modified. niques, namely the propeller shaft a. bearing overload, causing dam- alignment. The propeller shaft system of a warship age to the bearings and shaft; (Figure 1) can generally be described as Although there are many articles avail- a slender, flexible shaft supported by five b. bearing underload, causing shaft able about specific techniques of propeller or more bearings which are attached to a vibration and whipping; shaft alignment1"4, perhaps one of the complex, flexible foundation. The system c. high shaft bending stresses, caus- easiest to understand has recently been must be capable of transmitting propeller ing fatigue failure of the shaft ; published by Vassilopoulos4. In his arti- thrust and torque, while preventing the cle, Vassilopoulos suggests that the termi-

BULL OEAR

•-GCARBOX -MAIN A-BRACKCT -INTERMEDIATE an PLUMMER BEARINGS BEARING A-BRACKCT BEABING BEARING

Figure 1. CPF Propulsion Shaft Alignment

22 MARITIMI F N(;INI I:RING JOURNAL d. gear-tooth misalignment resulting Methods of Alignment from propeller shaft forces and causing possible tooth failure and There are several methods available noise; and for measuring or checking the state of alignment of a shaft system at any phase SHAFT e. stern tube seal leakage due to of installation and these are discussed in DESIGN large relative movement between detail elsewhere1"4. For the purpose of the hull and shaft. this paper, only a list of these techniques In the general case (and specifically for is necessary; to wit: CPF), it is necessary that the alignment a. piano wire — the simplest and \ plan be chosen early in the equipment most common method which is design stage so as to have the maximum still used extensively by ship- flexibility in modifying the design. When yards, basically a line-of-sight ALIGNMENT selecting an alignment procedure, the first method based upon a fixed SELECTION assumption is that all bearings will be reference; located along a line of sight, evenly b. optical techniques — similar to spaced and supported by a rigid structure. the above method, but using tele- A detailed computer analysis of the shaft- scopes or laser instruments, hence \ line is then used to calculate the bearing eliminating errors associated with reactions and shaft deflections. With these wire sag; results it is then possible to review the ALIGNMENT shaft system design and consider modifi- c. gap and sag — (generally used in IMPLEMENTATION <• cations to bearing location. If the bearing combination with a. or b.) con- foundation design has been frozen, then it sists of measuring closing toler- will only be possible to modify the bear- ances (at couplings) just before ing position either vertically or athwart- components are assembled; ships. This is accomplished by specifying, d. bearing reactions — the "nor- \—1 relative to the line of sight, the centre off- mal" approach in alignment set and inclination of the bearing in both implementation using hydraulic ALIGNMENT ^^^ the vertical and athwartship planes. It will jacks or load cells (this is used MONITORING be demonstrated later how this philosophy after a., b. and c. and is the pri- was applied to the CPF. mary proof that alignment has Having revised the bearing locations, been achieved); and Figure 2. further computer analysis is undertaken to e. strain gauges — a "new" tech- j Shaft Alignment Phases determine the impact of hull movement on nique in which gauges are affixed the shafting system. This calculation will to the shaft itself to allow for in- only provide a rough estimate for bearing service alignment checks (may be undertaken. The design alignment reactions. With the complexity of hull used with an RF transmitter to shall be achieved with the ship sub- structures and the limitation of present give dynamic strains with shaft in jected to low ambient temperature computer programs, it is not possible to motion). gradients and with the propulsion calculate the precise hull deflection. In the case of the CPF, the prime con- machinery in the warm running con- Thus, practical experience must be used tractor Saint John Shipbuilding Limited dition. " more extensively if modification of the (SJSL) has used a combination of the first Futhermore, the contract specifies the bearing positions is being considered at four methods. This is considered "com- this stage in the design process. position of the shaft axes relative to the mon" shipbuilding practice as these tech- bearings, gap and sag allowed and the Once the design has been established, niques were employed during bearing load and influence numbers. In the alignment process continues with the construction of the DDH-280 destroyers. order to achieve all of the criteria, the selection, implementation and monitoring It is the responsibility of the Prime Con- contract describes a "minimum" pro- phases illustrated in Figure 2 tractor (SJSL) to devise an alignment cedure to be followed. As previously (Vassilopoulos4). The selection phase procedure to demonstrate that he has met stated, this procedure should be subject lasts through the course of the contract the contractual requirements of the techni- to change during the design process (as and detailed design stages of the ship. It cal specification. has been the case). Specifically, SJSL is carried out by the ship's designer, ship- and their sub-contractor, YARD Ltd of yard specialist and occasionally by equip- CPF Contract Requirements Glasgow, developed a "Unique Pro- ment manufacturers. The implementation The requirement for alignment of the cedure" for the alignment and instal- phase is carried out by the shipyard during propulsion machinery is as stated in the lation6 based upon the original contract construction and the monitoring phase prime contract5. document.This is a "living" document commences during pre-delivery trials which will be constantly reviewed and and, in some cases, could continue "The design alignment of the propulsion machinery shall be achieved with revised by the contractor and DND until throughout the life of the ship with suita- the design alignment is achieved. The ble instrumentation. the ship afloat in a load condition approximating that under which the final stage of the process will be the trial Contractor's Sea Trials shall be of the installed system7 just prior to the commencement of sea trials.

JANUARY 1989 23 34 25.5

FWol DATUM

OPTICAL ( PROP. <£) INSTRUMENT 1. Erect bottom hull units. Establish datums. Optimize design shaft centre lines to ship seatings. 2. Ship gearbox, gas turbine raft, diesel engine and main thrust blocks.

3. Weld hull up to No. 1 deck. Temporarily install A-brackets. Drop stern. 4. Transfer aft datums to hull. Check propeller tip clearance. Align and weld A-brackets. 5. Align gearbox to design shaft centre lines.

6. Erect upper hull units. Bore A-brackets. Finalize design shaft centre lines relative to gearbox output shafts. 7. Align and bolt gas turbine raft to gearbox. Align gas turbines, diesel engine and main thrust blocks. Fit resilient mounts under rafts. 8. Align plummer bearings. Install shafting. Check pre-afloat shaft alignment and bearing loads. 9. Check afloat alignment. Realign inboard shaft bearings as necessary. Fit main shaft flexible couplings. Check afloat bearing loads.

Figure 3. CPF Alignment Procedure

CPF Alignment Procedure The exact details required for each applied to a physical environment. They easy to model. However, the difficulty step of the procedure, and most of the appear to be an exact method of determin- lies in determining exactly how to account reasoning behind those requirements, ing if a bearing or piece of machinery is for the artificial constraints placed upon a are discussed at length in the Unique aligned. What then is the mystery? Why ship when it is in the graving dock and not Procedure6 and will not be repeated is alignment a problem? supported by buoyant forces. here. However, Figure 3 is extracted The problem arises from the fact that from the reference to illustrate the steps It is also noted that because of the a ship afloat is not in a static environment. effects of creep and temperature distor- being followed. It is subject to hogging and sagging along tion upon the design, alignment is not nor- Discussion its length and, to a certain extent, racking. mally achieved by a straight line-of-sight The designer must allow for this when he process. This is illustrated in Figure I The methods used by shipbuilders to establishes his theoretical axes for align- which shows that the design shaft centre apply the alignment procedure are all bas- ment. This procedure is also relatively ically scientific measurement techniques

24 MARITIME ENGINEERING JOURNAL line for the CPF can be established opti- Limited. Also, the assistance of LCdr P..I. cally, however, the deflected shaft axis in Horsted, RN for suggestions and discus- the operational afloat condition requires sions about the "black art" was much that the barrels of the intermediate and appreciated. main A-brackets be slope bored. Also note the requirement for the aft plummer bearing to be raised 4.0 mm above the design shaft centre line.

When building a ship, the approach is References from the other extreme. It is almost impossible to predict what a ship will do 1. C. Archer and O.K. Martyn, Static and Dynamic Alignment, Trans I Mar E (C) when it is floated-up. Although breakage Vol. 91, Conf No. 4, Paper C31, readings and stern-drops are done, the pp. 24-31, May 1979; best a ship designer can hope for is that the ship will assume a fair curve upon 2. A.W. Forrest Jr. and R.F. Labasky, Shaft Alignment Using Strain Gauges, Marine float-up. Ships of the same class have Technology, Vol. 18 No. 3, July 1981, been known to move in completely oppo- pp. 276-284; site ways. The result is that the design alignment is such that it is a median about 3. Y. Latron et al, The Why and How of Shaft Intermediate A-bracket showing Alignment, Trans I Mar E (C) Vol. 91, installation of bearing stave. which there is a great deal of movement. Conf No. 4, Paper C29, pp. 3-11, May The ' 'black art" is to design a shaftline 1979; with enough flexibility to take into 4. Vassilopoulos, Static and Underway Align- account all of the possible configurations. ment of Main Propulsion Shaft Systems, This is usually a result of a great deal of NEJ, May 1988, pp. 101-116; experience and some good luck. That is 5. CPF Contract Schedule A, Part 3, Section why the procedures tend to be conserva- 3, Annex 2, V3B4P2S5 Propulsion tive and somewhat redundant in that in the Machinery Installation and Alignment; case of a design error there will be enough 6. CPF Propulsion Machinery Installation and information available to rectify the prob- Alignment Procedure Rev. 2, CDRL lem at minimum cost. In the case of the 368X, DID QA-015 04 May 1988; CPF procedure, it is felt that the inherent 7. Inspection of Installed Shafting, Gearbox flexibility of the shaftline and the very and Engine Alignment, Rev. SI, Trial basic approach should allow for correct Serial No. 619, Trial No. alignment despite the fact that this system M/24/150/201/00/000, 29 Oct 1987, is one of the first dual-shaft, cross- CDRL 371X, DID QA-019. connected installations in North America (albeit the DDH-280 raft is very similar). Conclusion The paper has tried to explain that the mystery surrounding alignment procedures is not so much one of technique of Setting up the boring machine for the starboard main A-bracket. achieving an "alignment" (the process is a controlled measuring process using any of several methods) as much as it is a problem of determining how a ship will behave while afloat. This can vary from ship to ship and hence the design alignment must have enough flexibility to allow for the dynamic environment that a ship must endure. In fact, for an MS officer, the procedures are precise and Prior to his posting last October, LCdr exact, however, the final result of a well- Staples served three and a half years as aligned machinery system is dependent the CPF Marine Systems and Quality upon the ability of the designer to predict Assurance Officer in Saint John, NB. He what his ship will do in a seaway. is currently with the directorate of nuclear safety policy at NDHQ. Acknowledgments The author would like to acknowledge that the majority of the CPF-01, HMCS Halifax pictures and figures are used with the permission of Saint John Shipbuilding Final check on the port main A-bracket bearing.

JANUARY 198") 25 Evolution of the Man/ Machine Boundary in Combat Systems

By Cdr Roger Cyr Introduction The Canadian navy has experienced quantum leaps in combat system technology over the past few decades. Much of this evolution has affected particular sensors and weapons which form a warship's combat suite, but significant technological advances have also been realized in the way in which sensors and weapons are integrated. And, this, it is believed, is the area that holds the greatest potential for technological advancement. The advent of artificial intelligence and mass data- storage devices is clearing the way for the development of so-called "smart" machines. Capable of high-speed, high- level decision-making, these smart machines will likely effect the gradual replacement of what is arguably the weakest element in the command and control of a combat system — the operator. Background In the 1960s the command and control (C2) function on board Canadian warships was performed solely in a manual Manual command and control of the 1950s fashion. Sensors and weapons were manually operated and monitored, tactical weapon assignment. But where the state machine in today's complicated combat data was transmitted via voice commands of the art in system design, processor environment. Indeed, the human may and tracks were plotted manually. Ten memory and processor speed placed cer- well have become the weak link in the years later, the new DDH-280 Tribal- tain limits on the extent of automation and process and his intervention may have to class destroyer brought forth a great integration in the ship, further automation be suppressed. improvement in shipboard command and of the CPF C2 process was undoubtedly control through some integration of the stifled somewhat by institutional con- There is plenty of evidence to prove various elements of the combat suite. straints and an unwillingness by humans that the human element is a risk factor in today's complex C2 environment. A Although there were marked ameliora- to accept too high a level of automated USN investigation into the downing of an tions in the Tribal's sensors and weapons decision-making. with the introduction of surface-to-air Iranian commercial airliner in the Persian missiles, the evolution in C2 technology Operator Shortcomings Gulf last July by the U.S. cruiser Vin- cennes revealed that the ship's Aegis air- was undoubtedly the decade's most sig- In many systems operator intervention defence system worked properly, but that nificant improvement in combat systems. is introduced in order to make up for defi- crew members misread the data which ciencies in system design. That is, the par- The 1980s saw the start of construction was presented to them. Under the stress ticular function could well have been of the new Canadian patrol frigates. of combat they expected to see an Iranian Again, these ships reflected major performed automatically, but the design F-14 fighter attacking their ship, so in did not cater to it and an operator function enhancements over the class of the previ- spite of the information being presented to was created to compensate. Many func- ous decade as a number of tactical func- them by the ship they assumed the aircraft tions still heavily dependent on human tions became automated to provide was an attacking F-14. Their psychologi- intervention, such as the identification of automatic detection and tracking and cal biases precluded them from making an limited capability in threat evaluation and threats, can best be performed by a

26 MARIT1MI 1 NCIM 1-RING JOURNAL Semi-automated command and control of the 1980s accurate assessment of the situation. time in a number of critical paths where to ignore the advice and fly the aircraft as According to U.S. psychologist Michael operator decisions are currently required. he read the situation. The aircraft Canter, an expert on perceptual bias, such crashed. human mistakes are almost unavoidable in Even in non-battle situations the conse- In each of these incidents had the air- battle. Where it is called human error, it quences of man attempting to cope with craft been totally computer controlled the is in fact human nature or human short- complex systems can be fatal. At an air proper evasive action would have been coming. show in Paris last June an A-320 Airbus crashed during a demonstration flight. A initiated automatically and both aircraft In other cases the risk factor stems sophisticated aircraft, the A-320 was would likely have survived. It was the from a human shortcoming much more designed to be able to prevent most pilot human element that placed the system at straightforward than anything so lofty as mistakes, but not quite all. The aircraft risk. The human himself, subject to the perceptual bias. It is the human's inability system prevents the pilot from taking the weaknesses of psychological and percep- to react quickly enough to a modern threat aircraft outside a safe flight envelope — tual prejudice, exhaustion, stress, indeci- at sea: USS Stark damaged by an Iraqi- overstressing the aircraft by turning too sion and a limited, volatile memory, that launched Exocet missile in the Gulf last tightly, or stalling by flying too slowly — was the weak link. May, HMS Sheffield destroyed by an but if he wants to, the pilot can still fly the The Future Argentinian Exocet missile in the Falk- aircraft into the ground. landsin 1982....The list goes on. A report From a technical point of view, the of the Stark incident stated outright that In Canada, recently, a CF-18 high- automation process has been limited only because missile attacks evolve so quickly, performance fighter was caught in a spin. by the memory space and data-transfer leaving little time in which to defeat even The on-board computer advised the pilot speeds of available processors. But tech- a single threat, computer-to-computer of the corrective manoeuvres to take, but, nical innovations of immense proportion automation should be used to cut reaction disoriented by the spin, the pilot elected

JANUARY l')89 ment in combat systems. These machines will be superknowledge-based systems which, through a logical process, will control all warfare functions during an engagement, ensuring optimum response to a threat by an elaborate deductive process. They will be able to model the environment based on procedures, doc- trine, tactics and rules of engagement, and search vast memory banks of parameters, looking for the optimum match to validate and verify the sensor and weapon data with a precision and speed that would be humanly unattainable. Conclusion The technological advances of the decades ahead are expected to bring unlimited innovations in the way in which war at sea will be conducted, particularly with respect to the man/machine boundary. It remains to be seen, however, if these innovations will be constrained by man's natural resistance to change and by his reluctance to accept machine-based decision-making processes.

Fully automated command and control for the year 2000, with the commanding References officer actually driving the ship with a joy stick. 1. Preston, A., The Smart Ship, Defence, Nov 87. are now taking place in computer hard- problems, make predictions, provide 2. Hill, B., Technology Failure or Human ware and data-storage devices. For exam- rationale and make decisions. The expert Error, Ottawa Citizen, 30 Jun 88. ple, where the pre-TRUMP DDH-280 system of the future will be able to per- 3. A Case of Human Error, Newsweek, 15 combat system memory space totalled form reasoning processes which will Aug 88. some 250 kilobytes, that for the new result in command and control systems no frigates will total over 15 megabytes, or 4. Navv Reports Raise New Questions About longer requiring human intelligence or Stark, News Supplement to IEE Spectrum, about 60 times as much. Logic and mem- intervention for decision-making. Jan 88. ory component density have been quad- rupling every three years (a trend The processing power to achieve these expected to continue) and processor mem- supersystems will be available in the very ory availability will be virtually limitless near future. What is still lacking is the with the advent of the four-megabit chip. software productivity to accomplish this with 100-percent reliability. However, it The development of efficient relational is expected that vast improvements in the processors will eventually ensure instant way we produce combat system software access to virtually infinite data bases, and will be forthcoming with innovative the optical disk will offer tremendous knowledge-based systems and better potential as a storage medium for tactical object-oriented programming languages data and images. As far as data transfer is such as Ada. concerned, great advancements are being The Smart Ship made with laser, electro-optical systems and Gallium Arsenide transmission and In the navy the aim for the future will reception circuits that are expected to be to produce the smart ship. The smart produce switching and driving speeds of frigate would see a reduction of crew size at least 600 megabits/second, or 60 times from the present complement of about 200 faster than the present SHINPADS. members to a maximum of 50 members. Commander Cyr is the DMCS 8 section Reductions in the operations branch head for naval computer technology at Expert systems are also now emerging would be achieved by the utilization of NDHQ. in military applications. These systems smart machines to replace the human as duplicate the kind of results achieved by the integrating and decision-making ele- human intelligence. They are able to solve

28 MARITIMI I NGINI I RING JOURNAL Looking Back: 1880-188 HMS Chary bdis 4- The Black Sheepf in Canada's |"Javal History '^""""••^w '

Could t exactly when, but soon winter gale in Saint JdHn ~more n Saint John Charybdis than a century ago delayed the cially raised by the Admiral from her moorings in a gale 1870s, the commander of the Canadian and damaged much of the shipping in the Naval Service Act of Canada by Militia came up with the idea or" drawing harbour. She had hardly been secured, almost thirty years ? The idea isn 't from the country's 90,000 fishermen and and the clamour of aggrieved shipowners as farfetched as it sounds. other seafarers to create a naval reserve. had not died down, when two Saint John He also suggested it would be of mutual citizens, attempting to go on board the benefit if the British government were to ship, fell through the rotting wood of her By LCdr Brian McCullough provide the Dominion with a warship gangplank and were drowned. That was You've probably seen a photograph of which could be used for coastal defence sufficient naval experience for the gov- Charybdis orphaned somewhere in the and naval training. ernment. The Admiralty was asked to take introductory pages of a naval history their gift back and in August 1882 the ship The Canadian government supported (Schull calls it a wreck at this point) was book. But did you ever wonder why this this recommendation and on October 8, towed to Halifax and turned over to appar- old three-masted steam corvette, the first 1880 the governor general sent a dispatch ently unwelcoming authorities of the warship ever owned by the Dominion to the colonial secretary stating that his Royal Navy. government, didn't rate a place in the government "would not be adverse to main text alongside the likes of Rainbow instituting a ship for training purposes if "Charybdis," Schull wrote, "became and Niobel the Imperial Government would provide a gruesome memory, a political flying Brief though it was, HMS Charybdis' the ship." The Admiralty responded by Dutchman which heaved over the horizon appearance on the Canadian naval scene offering HMS Charybdis, a decrepit when any naval proposal was advanced in 1881 -82 was to be long remembered as steam corvette just then limping during the next thirty years." That was an unpleasant episode in the country's homeward after seven years on the China her legacy. And it wasn't until 1909, in move toward naval self-reliance. Virtu- Station. She was offered at first as a loan, the face of a rapidly developing maritime ally overnight, what was supposed to have then shortly afterwards as a gift. The threat from Germany, that the cries of been the flagship of Canada's naval effort Canadian government accepted and sent " Charybdis!" began to be drowned out became instead a wooden-hulled pariah Captain Scott, a retired RN officer, to by the more strident, urgent calls for a despised by the politicians. For almost England to fetch her. naval service. On May 4, 1910, the Naval thirty years afterward, in fact, the cry of Things got off to a slow start when the Service Bill of 1909 was enacted and the "Remember Charybdis I" would be ship's chief engineer reported that the Canadian Navy came into being. A year enough to scuttle any naval proposal of boilers would not stand a winter Atlantic later, by command of the King, the serv- the day. crossing. The boilers were replaced at the ice was designated the Royal Canadian Navy. So what happened? Well, the story as expense of the Canadian government, and it's told in the Naval Service of Canada early in 1881 Scott "coaxed and coddled" and in Joseph Schull's The Far Distant Charybdis safely across the Atlantic to Ships goes something like this: Saint John, New Brunswick.

JANUARY 1989 29 News Briefs

Captain Garneau to promote manned space program

Four and a half years after becoming the first Canadian astronaut to fly a mission in space, Captain(N) Marc Gar- neau has retired from the navy to become deputy project manager of the Canadian astronaut program. A naval combat systems engineer, Captain Garneau flew on the Oct 5,1984 mission of the space shuttle Challenger. Even though he stayed on with the Na- tional Research Council's astronaut program in the years following his spaceflight, he said his decision to leave the navy last January after 23 years' service was not easy. "It's a decision I've been mulling over during the past year," he said. "I want to promote Canada's manned space program," the 40-year-old Quebec City native said. At the moment he is involved with solar radiation and shuttle luminescence experiments which will be carried into space on a future mission. Commodore Broughton (DGMEM) with Captain Garneau at the farewell. "I'm helping to get these experiments ready from an engineering and procedures point of view," he said. He has already developed the special graphics software which will be used with the Space Vision System during shuttle space-docking and Canadarm operations. "I would love to get back into space," he said," but I'm very happy to be in the support role. I would like to see other Canadians fly - -1 don't want Canadians to think it was just a one-mission affair and now it's over." Working on Canada's manned space program is "extremely exciting," Gar- neau said. "It's on the edge. I get the chance occasionally to roll up my sleeves....to do the kind of things I en- joyed in the navy - - a combination of desk work and field work."

Commander Dave Jacobson introduced the world's first "Marc Garneau" doll.

30 MARITIMI I NGINI 1 RING JOURNAL MARE selected to make CSME student-lecture tour LCdr Rick Francki, DMEE 2 manager of the DDH-280 cruise engine replace- Go ahead, string us a line. ment project, has been selected to repre- sent DND in support of the Canadian * Project updates Society for Mechanical Engineering's sixth annual lecture tour of Canadian * Special events engineering universities starting Janu- * People in the news ary 25. The tour is sponsored annually by the Chances are you 'II be CSME to bring university students up to date on topics which have a "significant the first to let us know. impact on Canada in the field of Mechani- cal Engineering." The Maritime Engineering Journal The DGMEM involvement this year DMEE, National Defence Headquarters comes at the Society's request for DND MGen George R. Pearkes Bldg., participation. LCdr Francki's presenta- Ottawa, Ontario K1A OK2 tion, "The Re-engining of the DDH-280s and Associated Engineering Problems," is being delivered as a case study of a current project involving hard mechanical engineering. There are 26 stops on the seven-week coast-to-coast tour, including one in King- ston in February for a combined presentation to students of the Royal Military College and Queen's University. Five French-language universities will also be visited before the tour winds up March 16 at McGill University in Montreal.

The Maritime Engineering Journal Salutes HMCS Assiniboine DDH-234 in service for Canada Aug. 15, 1956 - Dec 14, 1988

31 Index to 1988 Articles

January April September The Salvage of MV Essi Silje and Nuclear-Propelled Submarines for The Diesel Inspection Program Lloyd's Arbitration Hearing Canada by LCdr Richard Sylvestre by Edwin S. Chan by Capt Simon MacDowall ASW Frigate Electrical Propulsion - The Oil and Coolant Condition Analy- An Introduction to Stirling Engines and The Way Ahead sis Program (OCCAP) as an Equipment their use in Submarines by W.A. Reinhardt and J.R. Storey Health Monitoring Technique by Lt(N) Richard Sylvestre The Product Work Breakdown System by Michael Davies System Design — The Missing Element — Blueprint for CPF Construction Fabrication of Submarine Valves in by Cdr Roger Cyr by LCdr Richard Payne Canada The CPF Combat System Test and New Frigates a Shipbuilding Success by Z. Joseph Podrebarac Trials Program Story SWATH Ships — A New Term in the by Lt(N) David MacDougall by LCdr(R) Brian McCullough Equation Requirements Influencing the Design of Software and the MARE by Cdr John N. Edkins Canadian Naval Gearing by Cdr Roger Cyr The Replacement Cruise Engine for the by O.K. Nicholson Looking Back: The Nipigon Bow DDH-280 Tribal Class Destroyer Profile: Gloria Jessup — Secretary to Fracture by Cdr D.J. Hurl DGMEM by Cdr John Edkins, Lt(N) Mark MARE Olympic Torch Runners by LCdr Brian McCullough Gray and Clyde Noseworthy by LCdr Brian McCullough Looking Back: DDH-280 Gearbox Looking Back: Canada's First Naval Assembly Board by Steve Dauphinee

32 MARITIME INGIN1 l.RINC JOURNAL