THE MAGNETIC FIELD OF A SHIP AND ITS NEUTRALIZATION BY COIL DEGAUSSING

By W. C. POTTS, M.Sc, F.Inst.P.*

(The paper was first received 29th December, 1945, and in revised form 13th February, 1946. // was read before THE INSTITUTION 5 th April, 1946.)

SUMMARYIn November, 1939, the order was given to degauss all ships by

coiling. This decision was taken after a coil had been tried on a shipat Portsmouth and proved effective against the original German mag-netic mine. Mass treatment of this kind immediately showed manydifficulties. In fact, study of the problem had merely begun and verymuch investigation was required before it was possible to make de-gaussing effective against the much more sensitive mines which werelaid from 1941 onwards. This paper surveys briefly the study of themagnetic fields of ships made during the war, and indicates howcoils were placed to neutralize them.

(1) INTRODUCTIONBritish degaussing on full-scale ships began at the end of

November, 1939. Up to that time the testing of coils for de-gaussing had been carried out only on models. Examinationof the mine mechanism showed that it was set to fire on an in-crease of vertical magnetic fields of about 50 milligauss (mG).The model results had shown that ships could be rendered fairlysafe against such a mine by fitting a coil completely around theship and passing an electric current through it in a directionwhich would set up a magnetic field to oppose that produced bythe ship. No time was lost, therefore, in making similar testson a real ship. These tests proved satisfactory, and as a resultthe Admiralty issued the order to coil all steel ships as quickly aspossible.

The problem at this stage had hardly become scientific. It wasmuch more in the nature of a feat of electric wiring. There weremany ships to be coiled and much cable to be provided. Inorder to supervise the project it was necessary to create a newAdmiralty Department under the newly appointed Superintendentof Demagnetization (S.D.G.). The organization grew quicklyand soon many ships were being fitted with coils, a peak figureof more than 200 ships a week being reached at one stage.

Most of the original coils, or "girdles" as they were called bythe Press, were slung either outboard or in the scuppers. Thecoil followed the shape of the ship and it was usually clipped inposition; sometimes it was merely lashed on with rope. Themain requirements were that there should be enough ampere-turns and that these should be switched on in the correct direc-tion. Within six months most ships had some sort of coils inaction and casualties were very much reduced.

Magnetic compensation produced by these methods could notbe more than approximate, but fortunately this was enoughbecause of the initial insensitivity of the mine. However, closeexamination of the recovered mines had shown that settings asfine as 15 mG were possible, and it was obvious that somethingmore than approximate degaussing would become necessary.Furthermore, the polarity of the mine could be reversed in orderto fire under ships which were over-degaussed. An intensivestudy of the magnetic fields of ships was urgently necessary.

(2) EARLY EXPERIMENTAL ORGANIZATIONOne of the first requirements was a quick means of measuring

the magnetic fields under the ships. This was necessary for twomain reasons:

• S.D.G. Department, Admiralty.

(a) To find out how effective the coils were and adjust thecurrent in them to provide the best compensation.

(b) To find out how better compensation could be provided.A simple form of measuring range was planned and a rangegroup was formed to supervise the laying of such ranges at mostof the main ports in the United Kingdom. These ranges weremanned by naval officers and newly recruited scientists whoreceived operating instructions from the range group.

The rapid flow of results from the ranges soon made it obviousthat a central organization would be needed to analyse them. Asa result, the Range Analysis Group was set up in PortsmouthDockyard. The group found it necessary to publish a DailyRange Report giving recommendations for the coil settings ofall ships. This report is still issued and has now passed its1 700th edition. Much of the credit for the early success ofdegaussing must go to this organization.

A Trials Section was also formed. The work of this sectionwas made difficult by the scarcity of ships and the fact thatproblems had to be solved by experiments on any ships that couldbe made available even if only for a short time. Eventually itbecame possible to obtain a vessel, H.M.T. Sawfly, which couldbe used for an extended period solely for degaussing trials.Much of the original work was carried out on this vessel.

(3) EARLY RESULTSFrom the first range results it was evident that a single coil

would not be enough for many ships. It so happened that thefirst ships which had been tested with coils had been built on anEast or West heading, so that they showed fairly good compensa-tion with a single horizontal coil right around the ship. Thiscoil was known as the main coil, or M-coil. Unfortunately manyships are built on slips which point North or South, and in conse-quence they become quite strongly magnetized longitudinally,acquiring Permanent Longitudinal Magnetization, or P.L.M.Results showed that for a North- or South-built ship the M-coilwould reduce the magnetic field under the middle of the shipapproximately to zero, but that there remained a large positivefield under one end and a large negative field under the otherend. These fields were in some cases large enough to render theeffect of the M-coil almost useless.

These points are illustrated in Figs. 1 and 2, where the fieldsare measured under the keel of the ship at a depth below thesurface equal to the beam of the ship. The component of fieldmeasured is always vertical. Curve (a) of Fig. 1 shows the keelsignature for a ship built on an East-West slip, and curve (b)shows the reduction by energizing a deck M-coil at the optimumcurrent. It should be noted that this Figure leaves out certainfields due to the heading of the ship, so that it rather exaggeratesthe protection given by a deck M-coil. Fig. 2 shows similarcurves for a ship built on a North heading. Curve (6) showsthat the ship is still dangerous with the M-coil switched on.

To remedy this, fresh coils were tested. These took the formof two coils of length about one-third of the ship's length, oneon the forecastle (F-coil) and the other on the quarterdeck(Q-coil). One of these was set to aid-the M-coil and the other

[488 ]

POTTS:] THE MAGNETIC FIELD OF A SHIP AND ITS NEUTRALIZATION BY COIL DEGAUSSING 489

:+90r

Stern

Fig. 1.—Keel signature of ship built on East-West heading: measure-ments taken in U.K. with vessel on East-West heading.

(a)No degaussing

Fig. 2.—Keel signature of ship built on North heading: measurementstaken in U.K. with vessel on East-West heading.

to oppose it. These were effective, to some extent, but couldhardly be considered satisfactory. It was obviously better, ifpossible, to produce a process which would remove P.L.M. fromships. This process, which became known as "deperming," isdiscussed at length in a separate paper.* However, during thesearch for a suitable process, many ships had F- and Q-coilsfitted to compensate P.L.M.

Some important general principles were soon established.The model laws were shown to hold reasonably well on realships. The changes of field produced by coils were propor-tional to the current in the coils, and hysteresis was negligible.Also, the permanent magnetism of the ship was not appreciablychanged by prolonged running with the coils on. Many day-to-day questions arose, mostly about the location of coils or theampere-turns required. There were several questions on thesubject of screening.

About March, 1940, it was shown, as a result of model tests,that D.G. coils could be fitted inside the ship without detrimentto the compensation and without the use of excessive power.Such coils were, of course, much more difficult to fit, but theywere permanent. In contrast, some of the deck coils did notsurvive even one voyage in rough weather. The internal coilscreated a fresh set of problems concerning the location of thecoils, and the question of screening was even more in evidence.For example, it was not immediately realized by all concernedthat fitting the coil inside a protective steel tube causes noscreening at all, whereas passing it behind a longitudinal bulk-head might have quite a noticeable effect.

About this time it had been shown that ships could be de-gaussed without fitting coils. The process was known as

* AYLIFFE, S. H.: "Processes applied to a Ship to alter its State of Magnetization,"see p. 508.

"wiping" and was simply a method of wiping the sides of theship with an energized cable in order to leave permanent mag-netization to oppose the normal magnetization of the ship. Theprocess is dealt with in another paper. It is mentioned heremerely to indicate the relief it afforded to the coiling section.Many ships which had insufficient dynamo power to energizecoils were wiped. In general, small coasters became wiped shipsand the larger ships remained coiled.

(4) STUDY OF SHIPS' FIELDS IN GENERALThe fact that the German magnetic mine was actuated by

vertical field probably had some influence in the choice of thevertical component for ships' field measurements, but this is inany case the easiest component to measure. Consequently verynearly all field measurements made in the study of degaussinghave been of vertical components only. Thus, all measuredfields referred to from now on are vertical, likewise those shownin the Figures.

A brief description of the various components of ships' mag-netism and average values for the magnitude of the fields are givenin another paper.* Some further idea of the shape and magni-tude of the fields produced under the ship by each componentcan be obtained by reference to the Figures. Curve (a) of Fig. 1shows a keel signature of the effect produced by vertical mag-netism (V.M.) alone at beam depth. Fig. 3 shows a keel signa-

Fig. 3.—Vertical field due to ship's I.L.M. Measurements madedirectly below keel of a ship in U.K. waters.

Curve corresponds to North heading.

ture of the effect produced by longitudinal magnetism (L.M.)alone. Fig. 4 shows an athwartship signature of the effect pro-duced by athwartships magnetism (A.M.) alone.

Portbeam

(A

\ i\• d '

fie]

ical

t

-15

-10

-5

\\-io\

-15 ^

Stb ^ *beam v ^

/

Fig. 4.—Distribution beneath midship section of vertical field fromvessel's I.A.M. Measurements made in U.K.

Curve gives distribution corresponding to East heading of ship.

It should be made perfectly clear that all these Figures are syn-thetic. When a ship is ranged over an open range, the resultingsignature contains all the effects added together. It is the taskof the analyst to separate the signature into its component parts.The methods used in analysis are straightforward, and noattempt will be made to describe them in this paper. The processwas very soon reduced to a routine for simple cases, but it stillremains complicated when applied to a multi-coiled ship, which

* PARNUM, D. H.: "Underwater Measurements of Magnetic Field," see p. 435.

490 POTTS: THE MAGNETIC FIELD OF A SHIP

has to be made as safe as possible, e.g. a minesweeper. A typicaldiagram of a range record is given in the paper by Parnum.*

It is best at this stage to consider briefly each component ofmagnetization.

(5) VERTICAL MAGNETIZATION (V.M.)The process of acquiring vertical magnetization begins from

the moment the keel is laid, and continues to be "riveted in," as itwere, until the ship is completed and launched. Even then it isnot ended, because further changes may be brought about duringfitting out. At first sight, the whole process would appear to berather haphazard, but in fact this is very far from the case. Theprocess is analogous to that of magnetizing a bar of iron byholding it vertically and tapping the. end with a hammer. How-ever, tapping the bar indefinitely will not cause a continuous in-crease of magnetization, because a stable state is eventuallyreached. This is because of the demagnetizing field. The sameprocess occurs during the magnetization of a ship. Fig. 5 shows

Earth'? vertical field ""'

S S111!S S S 5

(Demagnetising field'R1

l l1111 .N N N N N N N

1 \ !

Fig. 5.—The approach to vertical equilibrium.The ship's vertical magnetism tends towards the state where '/?' equals 'Z'.

a cross-section of a ship bathed in an ambient field Z. This fieldproduces North poles in the keel and South poles in the deck.These induced poles set up a field R inside the ship. Now,clearly R must oppose Z, and equally clearly R can never exceedZ. In practice, magnetization proceeds until R is equal to Zand then ceases, provided there is sufficient vibration to reachthis state. This is called the equilibrium state, and the equili-brium magnetization thus achieved depends only upon the shapeof the specimen and the value of Z. Permeability of materialand thickness of plate do not matter in this condition.

It appears from observation that the vibration during buildingis enough to create the vertical equilibrium state in most ships,or very nearly so, along most of the ship's length. Sometimesthe ends of the ship have not quite reached it, but amidships thisstate is nearly always acquired. Now this principle is of funda-mental significance in the study of degaussing. Because of it,one knows in advance approximately how much vertical mag-netization a ship will acquire, and consequently it becomes pos-sible to predict the number of ampere-turns it will need in itsM-coil. But one can go much farther. In the equilibrium state,the magnetization depends upon shape only, for a given verticalfield, so that one might expect that sister ships built on the sameslip would have identical magnetic fields. Observation showsthat this is approximately correct, the discrepancies being ex-plained by odd structures which may be mounted in a piece suchas a mast or a funnel, and have acquired permanent magnetismduring separate assembly. This is very convenient from thedegaussing point of view, because it makes it possible to designthe same set of coils for sister ships. It also means that a scale

• loc. cit.. Fig. 2.

model will reproduce the fields of the real ship so long as it isshaken into equilibrium. Thus models can be used for manypurposes where full-scale investigation would be impossible.

As stated in the paper by Parnum, it has been found that theequilibrium vertical magnetization of a ship built in Englandwill produce on the average a vertical field of approximately90 mG under the keel amidships at beam depth. The value varieswith classes of ships because of the variety of dimensions. Forexample, the height of a battleship is about half the beam,whereas the height of the liner Queen Mary is almost equal tothe beam. Also the length of a trawler is about six times thebeam, whereas the length of a destroyer is about ten times thebeam.

If the earth's vertical field could be suddenly reduced to zero,the induced vertical magnetization (I.V.M.) of the ship wouldvanish and the permanent magnetism (P.V.M.) would remain.Normally, these two parts cannot be separated, but world-widedegaussing made it necessary to study them separately. Thisactually became a major problem, and it will be discussed in alater Section.

(6) LONGITUDINAL MAGNETIZATION (L.M.)During the building of a ship there will be a horizontal com-

ponent of the earth's magnetic field along the longitudinal axisof the ship, unless it is built in a direction which is precisely Eastor West magnetic. The vibration of building will thereforecause permanent longitudinal magnetization (P.L.M.). Inci-dentally, because most of Britain's main rivers run East or West,most of the "building slips lie North or South. ConsequentlyP.L.M. is very prevalent in British-built ships. However, aswill be shown in Ayliffe's paper, deperming is very successful inremoving P.L.M., so that it can be omitted from further con-sideration here. It is interesting to note, however, that becauseof the small demagnetizing factor longitudinally, equilibriummagnetization in this direction is not even approached.

When a ship sails and changes its course it is subjected to achanging longitudinal field. Consequently there is a varyingamount of I.L.M. which depends upon course. This meansthat, even if the M-coil compensates the V.M. perfectly, therewill still be field under the ship due to I.L.M. This usuallyamounts to about 15 mG in beam depth in U.K. waters. Sincethis is quite small compared with the effect of V.M., it can beignored in simple degaussing; but for accurate degaussing itcannot be neglected. Compensation of this field involveschanging the coil currents with heading and presents a muchmore difficult problem than M-coil degaussing.

(7) ATHWARTSHIP MAGNETIZATION (A.M.)A ship which is built on a slip pointing East or West has the

earth's horizontal component passing through it athwartships.It thus acquires P.A.M. during building. However, the amountis usually quite small because of the large demagnetizing effectin this direction. In the few cases where an appreciable amountof P.A.M. has been found, it has been removed by a processknown as "athwartship deperming."

Whereas the P.A.M. need not be considered further, the I.A.M.is important since all ships show the effect on courses near Eastor West. The magnitude of the vertical field produced underthe ship by I.A.M. is of the same order as that produced byI.L.M., i.e. 15 mG in beam depth in U.K. waters, but in this casethe fields are produced under the sides of the ship, with oppositesigns on the two beams. This field also changes with courseand involves changes of current in the coils when heading ischanged. It can be neglected in simple degaussing, but it mustbe considered when accurate compensation is needed.

AND ITS NEUTRALIZATION BY COIL DEGAUSSING 491

(8) THE M-COILThe M-coil is used for the compensation of V.M. As indicated

earlier, the original coils were||ther outboard or on deck. Theywere unseaworthy and liable to damage from many causes, butthey were also shown to be unsatisfactory magnetically. Earlyrangings showed that when the coil current was set correctly foramidships compensation, there was nearly always over-compensa-tion at bow and stern. This is shown slightly in curve (b) ofFig. 1, but the effect was much more evident in bluff-ended ships.It was clear that a single-deck M-coil could not compensate theV.M. accurately along the whole of the ship's length. This hadactually been anticipated by the model workers, who had pro-posed M-, F- and Q-coils for the compensation of V.M. as earlyas October, 1939.

Deck coils showed a much more troublesome fault than this,however. Many destroyers would have their M-coil currentsadjusted by magnetometer measurements while docked in about30 ft of water. When these ships ranged later in 50 ft they wereinvariably very much over-compensated. The cause of this wassoon evident: it was because the coils were too high in the ship,so that the depth law for the coils operated from a positionhigher than the magnetic centre of the ship. Since the mine thenin use operated on an increase of field, this over-compensationwas a good fault. But the introduction of a new mine firing ondecrease of field was expected almost daily, and this would makeover-compensation dangerous.

The first internal coils were tested in H.M.S. Viceroy atPortsmouth in May, 1940, and particular attention was paid tothe problem of correct coil height. The indication was that coilsplaced about two-thirds of the way up the ship's side would beat the best level. From that time onwards, permanent internalcoils were always fitted when time permitted, but it took sometime before results could be examined in any quantity. Then itwas evident that internal coils showed faults which were thereverse of deck-coil faults. They generally showed under-compensation at bow and stern because of the great reduction incoil area. This is illustrated in Fig. 6. Also, when fitted too

!.s-10

Bow Stem

Fig. 6.—Compensation of ship's vertical magnetism by an internalM-coil. The curves refer to the vertical field directly belowthe keel.

low in the ship, they showed general under-compensation atdepths greater than that for which the M-coil was correctly set.This study introduced the term "beaminess," which is illustratedin Fig. 7. The curves are all athwartship signatures of verticalfield in beam depth amidships. One curve is for the ship alone;the others show the effect of three different M-coils. It can beseen that the deck coil at A causes negative field abeam, and thelow coil at C causes positive field abeam. The coil at B is at thecorrect height and compensates correctly both abeam and underthe ship.

The problem of finding the best position to fit the M-coil hadfinally to be solved with models. Full-scale tests were obviouslytoo lengthy and too difficult. A full account of these researchescannot be given here, but, briefly, it was found that the correctheight for an M-coil depends on the ship, and that the optimumlevel is different in different parts of the ship. For example, thebest level amidships may be two-thirds of the way up the ship'sside, but only half-way up near bow and stern. Furthermore,

v! coil 'it correct level

Fig. 7.—Effect of height of M-coil upon compensation of vertical fieldfrom ship's vertical magnetism.

Measurements were made below a ship in U.K. waters.

the ampere-turns required are not the same along the whole lengthof the ship. This investigation led eventually to the 3-partM-coil, which is probably the best simple type that can be pro-duced. The main features are three separate parts, each em-bracing about one-third of the ship, known as MF-, MM- andMQ-coils. Each part is run at the optimum height for its ownpart of the ship, and the currents are separately controlled.Where necessary, there are re-entrant loops at bow and sternto prevent under-compensation. In the bows it is usually neces-sary, especially in fine-lined ships, to raise the coil above theoptimum level in order to get sufficient effective area. Fig. 8

M.M. M.Q..

Fig. 8.—Coil route for three-part M-coil.

shows the layout of a 3-part M-coil for a destroyer. These coilshave to be designed specially for the class of ship considered,and they have proved extremely effective in the accurate com-pensation of V.M.

It has already been mentioned that the original enemy mag-netic mine had a sensitivity of 40-50 mG, with a possibility ofsettings as fine as 15 mG. It should be mentioned now that,by combining acoustic and magnetic units, the enemy wereable to lay new mines having a sensitivity of 5 mG, in October,1941. So it is apparent that the original approximate degaussingwas then no longer adequate and that magnetic minesweepers,in particular, required the most accurate degaussing possible.

492 POTTS: THE MAGNETIC FIELD OF A SHIP

(9) F- AND Q-COELSF- and Q-coils have been used for several purposes. Their use

for compensating P.L.M. has already been mentioned. For thispurpose they are generally called FP, QP-coils. They have alsobeen used for trimming an M-coil, in which case they were calledFM, QM-coils. However, in modern systems they are used onlyfor course-correction, i.e. for the compensation of I.L.M., some-times called FI, Ql-coils, and that is the only use which will nowbe discussed.

It should be observed at the outset that the use of flat horizontalcoils for the compensation of longitudinal magnetization isobviously wrong in principle. One should use a longitudinalsolenoid or even longitudinal magnets. Both these methods hadbeen tested successfully as long ago as 1940. Unfortunately thedifficulties of fitting such devices were so great that they couldnot be considered for general use. Some U.S. battleships andminesweepers were fitted with solenoids, but the use of solenoidsnever became general. F- and Q-coils are therefore a com-promise. They are wound in a horizontal plane, one at eachend of the ship, and are energized in opposite directions in aneffort to compensate the vertical field produced under the shipby the I.L.M. Directly under the ship they compensate quiteeffectively over a limited range of depths, but they cannot beexpected to compensate correctly either abeam or at all depthsunder the ship because of the error in principle.

The history of F- and Q-coil design is very similar to that ofM-coil design. Full-scale work gave broad indications of whatwas needed, but models had to be used to produce the best designquickly. One simple principle emerges for these coils—theymust be as high in the ship as possible in order to achieve thebest compensation over a useful range of depths. The field dueto the I.L.M. poles diminishes with depth approximately accord-ing to an inverse 3/2 power law, whereas the law for the coils isroughly inverse square.

The best shape for F- and Q-coils depends entirely on the shapeof the ship. A bluff-ended ship, such as a tanker, concentratesits I.L.M. peaks right at its ends and requires short coils with re-entrant loops. On the other hand, a long tapering ship, suchas an aircraft carrier, has its I.L.M. effects spread over a largearea and requires much longer F- and Q-coils, usually withoutre-entrant loops. Such coils are illustrated in Fig. 9.

S55*J

\

Fig. 9.—Design of F- and Q-coils.Note that the coils are as high as possible and that in the case of tankers compara-

tively short length coils with re-entrant loops are required, whereas for aircraft carriersthe coils are long and without re-entrant loops.

The main difficulty with course-correction is the need to altercurrent with course. The maximum I.L.M. effect is on Northto South courses and is zero for East or West courses. Forcourses in between, the effect follows a cosine rule. It follows,therefore, that some member of the crew must be available toalter current with course. It also follows that there must be alimit to the number of current alterations that can be made.For this purpose only three course settings used to be given,

namely zero current between N 70° E to S 70° E and N 70° Wto S 70° W; two-thirds of full current positive between N 70° Eand N 70° W; and two-thirds ofiiull current negative betweenS 70° E and S 70° W. By this means the compensation wouldnot err by more than one-third of the total and would err byabout one-sixth on the average.

There was always the risk that course-correction might bewrongly used, especially during rapid manoeuvre, thus makingthe ship more dangerous rather than less. For this reason itwas difficult to apply in small ships where there were not enoughpersonnel to ensure correct use. Indeed, this practical considera-tion actually limited its use. However, the introduction of auto-matic course-correction has removed these objections. It isnow possible to couple a device to the gyro-compass so that theF- and Q-coil currents are automatically corrected within narrowlimits for any course. Mistakes due to misapplication have thusbeen largely eliminated.

(10) A-COILSThe A-coil is used to compensate I.A.M., and it can be used

also to cope with any P.A.M. which might exist. Correction ofI.A.M. comes under the heading of course-correction, but it ismore effective than tke compensation of I.L.M. because it ispossible to provide a coil which produces the right component offield to do it. The A-coil is usually fitted in the form of twovertical-plane coils run longitudinally through the ship. There isone coil each side of the centre-line roughly half-way betweenthe centre-line and the ship's side. The top limb of each coilis high in the ship, and the bottom limb is usually as low as pos-sible amidships. The coils taper towards bow and stern, andstop short some distance from the ends of the ship. Fig. 10shows the run of the A-coil in a battleship.

Fig. 10.—Design of an A-coil for a battleship.

Magnetically, an A-coil designed from a model is very nearlyperfect. It is by far the most successful of all D.G. coils.Compensation is correct for all depths, and abeam too. Un-fortunately, these coils are extremely difficult to fit, and this haslimited their use to the most important ships. It is desirable tobuild the coil into the ship during construction, and this hasusually been the procedure adopted.

The usefulness of F- and Q-coils is very much reduced unlessan A-coil is also fitted. This has been fully realized since 1941,and many ships have gone without F- and Q-coils because of thedifficulty of fitting an A-coil. It is worth remarking that at theend of the war the Germans had just realized that compensationof I.A.M. was necessary. This had been overlooked becausethey adopted loop ranges for their measurements, and loops donot respond to I.A.M. because there are equal contributions ofpositive and negative flux passing through the loop at the sametime.

The remarks on automatic course-correction as applied to F-and Q-coils apply equally to A-coils.

AND ITS NEUTRALIZATION BY COIL DEGAUSSING 493

(11) MODERN D.G. EQUIPMENTA fleet minesweeper is a type of ship which is fitted with the

best D.G. available. In 1945 it had six carefully designed D.G.coils, consisting of a 3-part M-coil and F-, Q- and A-coils,The course correction is automatic. The ship is carefully de-permed somewhere near its building yard, and it then proceedsover the nearest D.G. open range, making several runs. Fromthe analysis of the range records, the ship is given the currentvalues for all its coils for anywhere in the world. Thereafter,only occasional checks are necessary. The ship is then safe mag-netically to a degree considered unlikely of achievement in 1940.Tt can proceed about its normal duties with only negligible riskfrom even the sensitive magnetic mines. Of course, no ship isabsolutely safe, and for shallow-water sweeping it is necessaryto use wooden ships. Even if the water is only 6 ft deep thereare suitably degaussed vessels available.

Magnetic mines have been used extensively against thiscountry throughout the whole of the war. The initial success ofour approximate degaussing of 1940 did not lead the enemy todispense with this weapon. On the contrary, it spurred him onto greater efforts, and he produced new types at regular intervals.His final effort was a combined pressure-magnetic unit, whichwas first employed during the Normandy invasion. In fact, alarge proportion of the mines used off the Normandy beach-heads were magnetic in principle, and careful analysis of .thecasualties and the mines swept has shown quite conclusivelythat careful and systematic degaussing saved very many of theships used in this operation.

Since it is quite impossible to cover all the problems asso-ciated with the study of coil degaussing in a single paper, it isproposed to present only a selection of the problems which havebeen considered.

(12) AMPERE-TURNS FORMULAWhen coiling first began, one of the most urgent problems was

to provide some rule of thumb for estimating how many ampere-turns would be needed for any given ship. It was important tohave enough, but overcoiling would have been extravagant andthere was little copper to spare. The first rule used was AT= 16/t, where h is the height of the ship in feet. This was soonshown to be inadequate, and it was increased to 24h. Even withthis value many ships were under-coiled. The subject is raised herebecause it provides a very simple example of the kind of statisticalwork which proved very useful in the study of degaussing.

It was necessary to decide on a formula that would coveradequately, say, 90% of the ships. The problem was thenreduced to the examination of the range records, as they becameavailable for any class of ship, to see how many ampere-turnswere actually needed when fitting new installations to ships ofthat class. These were expressed in terms of the height of theship, though the beam would have served equally well, andentered on a suitable diagram as the work proceeded. Eventu-ally, there were enough results to reach a decision and a valuecould be taken from the diagram to cover 70%, 80% or 90% ofthe ships. Fig. 11 shows a histogram compiled for internalcoils in Empire ships. The ordinate represents the number ofships in each group, and the abscissa gives ampere-turns interms of h, divided into suitable groups of 2h. The vertical linethrough 35h has 80% of the ships on the left. Thus, 35/*ampere-turns will be adequate for 80% of Empire ships, andadoption of this formula will leave only 20% undercoiled.

(13) THE EFFECT OF LATITUDE ON VERTICALMAGNETIZATION

So far in this paper, for the sake of simplicity, considerationof a ship's magnetic field has been limited to one latitude.

I35A

io

28 30 32 54 36 58M coil setting expressed as multiple of ship's height

Fig, 11.—M-coil settings required in U.K. waters by a group ofstandard Lithgow merchant ships.

The histogram shows the number of vessels having specified ampere-turn settings.

Actually, the principal mining menace was in the Nore Command,which is almost small enough to be considered one latitude.Unfortunately, ships do not remain in one latitude, nor did theenemy limit his magnetic mining to the Nore Command. Mag-netic mines were laid in many areas, including the Suez Canal.

The vertical component of the earth's magnetic field (Z)varies over the navigable seas from +0-60 gauss in the Arcticto —0-60 gauss in the Antarctic. It varies over the BritishIsles alone from +0-43 gauss on the South Coast to+ 0-47 gauss at Scapa, i.e. nearly 9%. The horizontal com-ponent (H) varies from zero at the magnetic poles to 0 • 40 gaussin the Gulf of Siam. (Scapa, 0-155 gauss; Isle of Wight,0-185 gauss.) Thus, as a ship moves about the world it issubject to changing fields, and the ship's magnetization isaltered accordingly. This raises numerous problems in thestudy of degaussing. It is obvious that D.G. coils will requiredifferent currents when the ship changes latitude, but the questionis much more complicated than that.

Consider a very simple case. A ship is built on the Clydewhere Z is +0-46 gauss, and presumably achieves equilibriumvertical magnetization during building. It passes over a D.G.range and its M-coil is set at 800 ampere-turns for the area ofthe range. The ship now sails to Fremantle, where Z is— 0-49 gauss. The M-coil setting it now requires will not bevery important if there are no magnetic mines at Fremantle.But suppose the ship now cruises for six months in Australianwaters and then returns to the Thames, through the most denselymined area; it may now require a different M-coil setting.

In order to decide whether a new setting is required, one mustfirst know how much of a ship's equilibrium vertical magnetiza-tion is induced and how much is permanent, and in what waythe "permanent" part varies. A simple experiment was triedon H.M.S. Venomous in a lock in Portsmouth Dockyard. Theobject was to remove the earth's vertical field in order to measurehow much V.M. was permanent. A cable was passed aroundthe lock and energized with sufficient current to neutralize theearth's vertical field on the ship. The ship's field was measuredwith and without the cable energized. Of course, this cannotbe expected to give an accurate answer, because the neutralizingfield is not uniform. However, it showed that roughly 40% ofthe V.M. vanishes when the ambient vertical field is removed,i.e. I.V.M. is 40% of the total V.M. Model tests confirmed thisresult, but it has very limited value, because in both cases theship was at rest and completely free from vibration. Clearlythe important thing to know is how rapidly the P.V.M. vanishesunder the shaking effect of sailing in a reduced field.

The only way to get this information is to measure the ship'sfield as it moves about the world* and this requires D.G. rangesat suitable ports abroad. The project was commenced late in

494 POTTS: THE MAGNETIC FIELD OF A SHIP

1940, and it was not until the spring of 1941 that there wereenough ranges abroad to produce an appreciable number ofresults. Each ship ranged at ports where D.G. ranges existed,and the results were sent to this country for examination. In thisway the analysts were able to follow the magnetic changes ofindividual ships as they proceeded from latitude to latitude.

This examination was essentially a very slow one, and resultswere long in coming. However, the homeward-bound shipscould now be assured of greater safety on arrival in home waters.Ships returning through the Mediterranean could range atGibraltar and get the correct current to use for the U.K.Similarly, ships returning via the Cape could range at Freetown.In addition, magnetic mines began to appear in the Mediterraneanand other areas, so that the ranges abroad had to cater for theirown region as well as to provide the most probable currentvalues for home waters.

In order to make possible the alteration of M-coil current tosuit different latitudes it was necessary to devise some simplesystem which could be understood and followed by ships'personnel. A chart of the world was provided on which therewere 12 numbered zones of latitude corresponding to the linesof equal vertical field. In merchant ships the M-coil wasprovided with a series of numbered studs or settings and theship was given the setting to use in each zone. This zoned M-coilsystem did not permit precise setting of the current for all areas,but it allowed accurate setting in home waters by means of aspecial resistance, and limited the errors elsewhere to quite smallproportions.

Meanwhile, the examination of P.V.M. continued, andeventually it appeared that the facts could be explained byassuming the equilibrium vertical magnetization to consist ofthree parts, namely:

1. Induced 50% . .

2. Soft permanent 20% .

3. Hard permanent 30% .

Changes immediately with changeofZ.

Exponential in decay with changehalf completed in about 30 days.

Exponential in decay with changehalf completed in about 4 years.

These values differ somewhat with different classes of ships; thefigures given are for merchant ships.

It is clear that the assumption of three sorts of magnetizationis somewhat arbitrary, because there must really be an infinitenumber of degrees of hardness. But statistically these assump-tions are useful and make it possible to predict what M-coilcurrents will be needed by a ship in any part of the world.

Early in 1943 it became possible to agree with the UnitedStates authorities upon a universal system of D.G. charts to beused by all ships, and this scheme was launched in July of thatyear. Thus it became possible for any ship, whatever its flag,to be treated by similar rules anywhere in the world. Thisagreement proved of great value and simplified the work ofBritish, Dominion and U.S. authorities.

(14) THE EFFECT OF LATITUDE ON LONGITUDINAL ANDATHWARTSHIP MAGNETIZATION

It has been shown that I.L.M. and I.A.M. effects are variablewith course and have peak values of about 15 mG in beam depthin U.K. waters. The effect also depends upon the earth'shorizontal component (H). Fortunately, it has been shown byexperience that the magnitude of the effect is directly proportionalto H. Thus one can say that, if the field produced in beamdepth is 15 mG where H is 0-17 gauss, it will be 30 mG in thesame depth where H is 0-34 gauss. This makes prediction ofcurrent values in F-, Q- and A-coils very simple, and a coursecorrection chart, similar to the M-coil chart, is in general use.

The values on this chart, once given, need no alteration evenafter prolonged sailing in southern latitudes.

The magnetic conditions at Singapore provide an interestingstudy in this connection. The values of the earth's field areZ = — 0-10gauss and H— 0-39gauss. Thus the verticalmagnetization of a ship in that area is very small, and only asmall value of M-coil current is needed. On the other hand,the I.L.M. and I.A.M. are relatively very large, and the currentsrequired in the F-, Q- and A-coils are more than twice as greatas those required in U.K. waters. Thus, at Singapore, coursecorrection is of much greater value than the M-coil, which isthe reverse of the conditions at home.

(15) MAGNETIC CARGOIt has been pointed out that the value of equilibrium vertical

magnetization depends only upon the shape of the ship con-sidered, and that the vertical field inside a ship in equilibriumwill be zero. Now this means that one can fill up the holds of amei chant ship with scrap iron without affecting the external fieldat all. This assumes that the random positioning of separatepieces of iron will be such that any magnetization they possesswill be neutralized in bulk. It was some time before carefulattention could be given to this problem, but eventually theappearance of some abnormal ship's signatures enforced closerstudy and disclosed one of the most interesting results in thestudy of the magnetic fields of ships.

One can reach two obvious conclusions about magnetic cargo.One is that the effects of internal D.G. coils should be enhancedby its presence, and the other that the I.L.M. will be increasedbecause longitudinal fields inside the ship are not zero. Boththese effects are shown in practice, but they do not cause muchtrouble. A third conclusion is also fairly obvious, namely that,if the magnetic cargo is systematically magnetized in any onedirection, the effect must show outside the ship. At first sightit would appear that such a disposition of the cargo inside theship would be unlikely, but events proved otherwise.

If steel rails pass through the rolling-mills in a North-Southdirection, all the conditions for strong magnetization are present.The rails are hot and are subject to great mechanical stress.Thus every rail will become a large permanent magnet. Nowit has been shown by investigation in several places that themethod of handling the rails from the mills right up to the loadingin a ship's hold can be so systematic that no rail ever getsturned round relative to its neighbours. Thus it is possible for,say, 1 000 tons of steel rails to be stowed longitudinally in ahold so that all the North poles are at one end and all the Southpoles at the other. When this happens, the magnetic fields pro-duced outside the ship by the cargo are very much larger thanthose produced by the ship itself. Furthermore, D.G. coils donot neutralize these fields (see Fig. 12). The only method ofpreventing this danger is to load the rails so that half are turnedround, but many ships actually sailed in such a state before thecause was traced.

The effect of such a cargo on the ship is interesting. The ship'svertical magnetization is stationary only in the equilibrium state.In any other condition it will tend to change with vibration.Now consider such a ship during a voyage. The various strainsdue to hogging and sagging, quite apart from engine vibration,all tend to take the ship back to its original state of magnetization.This can be achieved in only two ways. Either the steel railsmust lose magnetization, or the ship's hull must acquire mag-netization of opposite polarity to the rails in order to neutralizetheir effect. There may be some of each, but obviously therewill be less change where the magnetization is more permanent,i.e. presumably the rails, considering how they were magnetized.It follows, therefore, that when the cargo is eventually discharged

AND ITS NEUTRALIZATION BY COIL DEGAUSSING 495

+120

3 +80

•a +40"55

a.C/3

Vessel laden with 1000tons of steel rails •

-40

Fig. 12.—The effect of steel cargo on the vertical field of a degaussedship. The measurements were taken directly below the keel.

the ship will exhibit strong signs of having a magnetic cargoof the same sort, but turned round the other way, i.e. a weakerimage of the cargo. This did, in fact, happen to many ships,and special treatment had to be given to remove the effects.

A further effect of magnetic cargo arose in wiped ships.Wiping will be considered at length in another paper,* but it willsuffice here to say that the treatment is designed to reduce thetotal vertical magnetization to zero. It was found that many

• AYLIFFE, S. H., loc.cit.

wiped ships, having ranged with a cargo of scrap iron aboard,were recommended for retreatment. After unloading the cargothey were found not to need retreatment. Once the problem isposed, the solution becomes fairly simple. If there is zerovertical magnetization, as in the wiped state, the earth's verticalfield will pass right down through the ship. (Only the equilibriumstate gives perfect screening.) Thus the scrap iron is exposed tothe earth's field and produces its own field under the ship. Thiswas carefully tested with models and found to be the correctexplanation.

(16) CONCLUSIONNo attempt has been made to cover everything which might

legitimately come within the scope of the paper. The chiefaim has been to give an account of the features thought to havethe most general interest. Consequently, many aspects of thework have been omitted.

There have been contributions from the Dominions, and muchof the work has been paralleled in the United States, with whomthere has been very close liaison and complete agreement onpolicy.

(17) ACKNOWLEDGMENTSAmong the many individuals responsible for the advance-

ments described in this paper, it is desired to make specialmention of Messrs. R. d'E. Atkinson, E. C. Bullard, S. Butter-worth, J. Hext-Lewes, H. R. Hulme, A. T. Pickles, R. R. Roscoe.

[This discussion on the above paper will be found on page522.]

VOL. 93, PART I. 30

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具

http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/vip.htmlhttp://www.xuebalib.com/db.phphttp://www.xuebalib.com/zixun/2014-08-15/44.htmlhttp://www.xuebalib.com/

The magnetic field of a ship and its neutralization by coil degaussing学霸图书馆link:学霸图书馆