REPORT No. 20 S
January 1955 sSTUDIECENTRUM T.N.Ö. VOOR SCHEEPSBOUW EN NAVIGATIE
AFDELÍNG SCHEEPSBOUW/ - NIEUWE 'LAAN 76 - DÈLFT(NETHERLANDS RESEARCH CENTRE T.NO. FÓR. SHIPBUILDING AND NAVIGATION) ('SHIPBUILDING DEPARTMENT - NIEUW'E LAAN 76 - DELFT)
*
An analysis öf the application
of aluminium alloys in ships' 'structures.
Suggestions about thç riveting between ste el
and aluminium alloy ships' structures.
by
PROF. IR. H. E. JAEGER.
Summary
A subdivision into groups, based upon the most noteworthy properties of aluminium alloys as used on board ship, is mâde. Fóllowing this subdivision, a description is given of the principles underlying the attempt to define the scantlings Qf aluminium-alloy superstructures and deckliouses. The.influence. of such tructures on the longitudinal strength of the ship: is stúdied.
A list of requirements for the scantlrngs of light-metal superstructures and deckhouses is given with a new proposal embracing the criticisms of the existing rules and regulations. An extensive bibliography follows.
§ 1. I,itroductión
This study must be viewed firstly as following that published as report No 3 of the Netherlands'
Research Centre T.N.O. for Shipbuilding and
Navigation, entitled: "Practical Possibilities of
Constructional Applications of Aluminium Alloys to Ship Construction" [1].
In the second place it seeks to serve as a step
towards the establishment of international regu lations for the standardization of the composition and the dimensions of light-metal alloys used in shipbuilding. This work has been entrusted to the
working group (W.G. 1) of the ISO/IC 8.
A combination of these two purposes is simple to achieve, as we have only to reconsider the two last paragraphs, 13 and 14, of the report mentioned,
and to rewrite them, whilst noting the differetit
regulations laid down by various organisations and subjecting them to technical criticism.
A few words about the desirability of using light-metal alloys are appropriate here. The application of aluminium-alloys to ships' structures depends on a complex of economic and technical consider-ations. A review of those parts of the ship which might be constructed in such alloys is only possible by an evaluation of these considerations.
The following subdivision is an attempt to give
an evaluation of the different considerations and is set out in such a way that a discussion of each separate type of construction mentioned is not
necessary. As a criterion of each group, the two
most characteristic properties of the aluminium-alloys are taken, viz, their great resistance to cor-rosion and their low specific weight. A subdivision into three groups is considered necessary, one group embracing those structures for which both
charac-teristic properties are important.
After the subdivisiàn into groups it will be
necessary to reconsider those groups which may be amenable to strength calculations: Several parts of the ships' structure require regulations for scant-lings, and a critical survey of what has been done
is given.
The method of determining these scantlings may be either comparative with the former steel
con-struction or the determination of the light-alloy
construction by means of a direct strength
calcu-lation, where only the mechanical properties of aluminium and its alloys replace those of steel
[10], [11].
A further point is the question of which criteria should be used as the basis for comparative calcu' lations. These may be given as:
a. The basis of equal strength as compared with
steel construction. This means that the factor of safety to the breaking point of the material, in
both cases (steel and aluminium-alloys), is at least
equal.
The basis of equal stiffness. In this case the deflection of both steel and aluminium-alloy struc-tures is the same.
A compromise between both preceding bases is, that the difference in deflection of the structural part in question is 50 % of that between the con-struction calculated on equal strength and the one calculated on equal stiffness.
In the following pages these three bases will be indicated as: basis of equal strength, basis of equal stiffness and basis of 50 % deflection.
2. Subdivision in groups of ships' structures,
which may be constructed in
aluminium-alloy with reference to the general advantages of such structures
Group I: Ships' structures made in light-metal alloys, where the advantages arise from the fact
that the specific weight of. these alloys is much lower than that of steel.
As already mentioned in [1], the decrease in weight of structures made in aluminium-alloy
instead of in steel can be used to:
Increase the carrying capacity (see A and G). Either to improve stability or, when: stability remains the same, to decrease the beam (see B, C, D and H)..
Decrease the resistance of the ship (see A and.G).
Decrease the dimensions of the ship with the same carrying capacity so that .the same speed can be attained with less power (see A and G).
Increase the radius of action with equivalent speed and bunker capacity (see A).
Reduce the draft, especially that of small ships, which assists navigation in shallow water
(see A sub 2).
Increase the depth of hull while maintaining displacement, so that with the same or a slightly
greater carrying capacity, a better cubic can be
obtained. As the centre of gravity i1l also rise, a decrease of beam will not be possible.
Muckle gives much consideration to this type of weight-saving by the use of light-metal alloys in an article in "The Shipbuilder" [.2]. From this the figures 1 and 2 are taken, giving direct information about sOme advantages obtained by weight-saving. Basing himself exclusively on this saving, Muckle
comes to certain conclusions about the possible reduction of the breadth, displacement, fuel-cost and capital-investment.
Indirect advantages of the saving in weight are better stress-distribution (see C) and greater ease in handling (see D).
The question of corrosion-tests will be treated by a special panel of ISO/TC 8/WG i and will be
treated in the Netherlands by
Corrosion-Com-mittee V.
Linings of fish-holds [14], [15], [16]
Covering of insulation in refrigerated ships
for the transport of fish, meat, vegetables. and fruit [17]
Covering of insulation in refrigerated spaces of passenger- and cargoships
Covering of oil-tanks
Covering of wine- and chemical-store tanks. Interior of galley's
Interior of kitchens
Interior of sanitary installations Port-holes and windows [18] Ships' furniture
Ships' decoration Electrical installation.
Group III: Ships' structures made in light-metal-alloys, where both the advantage of light-weight
and high resistance against corrosion are of
im-portance.
Under G those parts are listed, where construc-tion in aluminium-alloy has a direct influence on
the carrying capacity and the propulsion. Under
H those parts are listed which have a particularly good effect on the stability of the ship.
Ventilation-ducts
Pipe-lines
Parts of machinery
Floats, buoys and skimming-boards for
fishing-vessels.
Lifeboats [12] Lifefloats
Funnels [1], [19], [20].
§ 3. Ships completely constructed in
aluminium-allo y
These structures are given in group l.A. The
possibility or the desirability of constructing a ship
totally in light-metal must be judged from an
economic point of view in the first place. A calcu-lation of earning power will show that thepossi-bilities are, in general, limited to small and fast
ships, like those mentioned in this group. General rules are very difficult to give and each case has to be calculated individually. The small size of
these ships makes the question easy from the tech-nical aspect, though comparisons with steel ships are not advocated. Indeed the conditions of loading are quite different, due to the large weight-savings. As the hull-weight is determined by the load, these ships must be calculated directly as structures in light-metal.
A. Complete structure in aluminium-alloy for: Very fast ships such as:
Naval craft Police-boats [3] Patrol-boats Yachts [4]
Small passenger-vessels for inland navigition Whale-catchers.
Ships navigating in shallow water:
Barges [5], [8] and [9]
Coastal lifeboats.
B. Masts and derricks [10], [11] Davits [12] Deck-auxiliaries Deck-ventilators Rails. C. Superstructures Deckhouses
D. Hatches and hatchways
Hatchbeams and fore- and afters [6], [7, [8] and [9]
Removable masts Removable derricks
Ladders, stairways, gangways [13] Deck-awnings.
Group II: Ships' structures made in
light-metal-alloys, where the advantages arise from the fact that the resistance to corrosion is considerably higher than that of steel.
The use of aluÍninium-alloys shows many ad-vantages for parts of the ship where longevity and low cost of maintenance are essential (see E and F), or where expensive materials can be replaced by these alloys (see F).
As already pointed out in [1] there are three
types of corrosion:
General corrosion, which causes the surface of the material to be entirely removed by chemical solution of the oxide-skin.
Local corrosion, pit corrosion or pitting. This
is always electrolytic, a result of local galvanic elements. The latter come into being through dif-ferences of potential between the components of the alloy, sea-water generally acting as the électro-lyte. The corrosion penetrates deeply over a small surface, causing the peculiar forming of pits; hence the name.
Intercrystalline corrosion, which takes place
by the corrosion of the metal along the "grain"-boundaries. As the grains of aluminium may be considerably larger than those of iron and steel, intergranular corrosion is much more likely to occur. The Al-Mg-alloys with a large amount of Mg are typical of this.. Metallurgically there is a remedy, but the practice is now to restrict the Mg-percentage to 4.7 %.
il
i
w w w > O< to
18
1.6w-I__
I_
I
-ddißr
:10o/68
h, HEIGHT 40 OF WEIGHT.GAIN ABOVE C.G. IN M. --5 678
IO Il ia 13 14 15 Jó h IN M.Fig. 1. Decrease of the height of ¡he centre of gravity o a ship in conjunction with the gain in weight (g) through the substitution of steel by aluminium-alloy at a height (h) above the C.G., (V) being the total displacement.
o
¿
s
+
POSSIBLE REDUCTION IN BREADTH OF SHIP IN M.Fig. 2. Influence of ¡be gaul in weight (g) through substitution of steel by aluminuom-alloy, on the ship's breadth, assunsing
constant-stability value.
As the modulus of elasticity of
aluminium-alloys (Young's modulus) is quite different fromthat of steel,
the buckling phenomena are notcomparable. Therefore,
not oniy the
bending-stresses, but also the buckling-stresses must be con-sidered separately. It is advantageous to use
longi-tudinal framing and the sheathing of the decks
with wood must be taken into consideration (see also § 8). Due to the lower resistance to buckling, the frame-spacing of aluminium-alloy constructed ships must be smaller [21].
Nevertheless, if comparative calculations are still made, Corlelt's and Muckie's methods and diagrams can be used [22] and [21]. These diagrams are also published in [1] as figures ;
io, 13, 14, 1, 16,
17 and 18.
The high price of the aluminium-alloys makes it desirablé to save weight by the use of welding. Modern welding-techniques make it possible to compete with riveting, which is still the normal
method of connecting light-metal plates and sec-tions of small thicknesses [23].
§ 4. Ships' fittings completely constructed in
aluminium-alloys
These structures belong to groups I.B. and I.D. The use of light-metal is governed by the wish to improve the ships' stability. The question of greater ease in handling is also a factor. This applies
es-pecially to parts which determine the scantlings of lifting gear. The cost of these parts is largely
influenced by easier fabrication in comparison with steel or other heavier metals.
As far as masts and derricks are concerned, calcu-lations against buckling or bending and buckling combined, make direct strength calculations for the
aluminium-alloy structure necessary [10], [11].
On the other hand, the stiffness of davits [12] and
stairways [13] must be the same as that of the
comparable steel structures and thus comparative calculations must be used. The same applies to ven-tilator-ducts, deck-rails, hatchbeams, etc. [6], [7]. A calculation based on equal bending stress
(equal strength) gives the requirement: Ja/ya = 1.65 la/ya. This factor of 1.65 is determined by applying the same safety factor against breaking
for the aluminium-alloy with 4 % Mg, using its
lowest breaking-point (26 kg/cm2) in comparison
to that of ordinary ships' steel, using its mean
breaking-point (42 kg/cm2). As E8/E5. = 3, where E8 = Young's modulus for steel and Ea = Young'smodulus of aluminium, the deflection becomes of the equivalent steel structure. Thus if fa is the
deflection of steel and fa that of the
aluminium-alloy
= 1.82 fa
For hatch-beams and fore- and afters in partic-ular, the necessary modulus must be found in the thickness of the web, otherwise wrinkling of the web may occur. Therefore the thickness of these webs, constructed in aluminium-alloys must be de-rived by the formula:
t, 1.44 t8 where
ta thickness of aluminium-alloy plate t.8 = thickness of steel plate.
For the determination of this formula and the calculation of hatch-beams with equal stiffness
(same deflection as steel beams) see § 8.
It is interesting to make hatch-beams for ships navigating in inland waterways in such a manner that they can float. This is easily possible by design-ing structures, which have the necessary
displace-ment [8], [9].
§ 5. Superstructures and deckhouses constructed
in aluminium-alloys. General observations. It is considered necessary to give first of all some definitions and axioms:
A long superstructure or a long deckhouse has
a length that is greater than 15 % of the ships'
length L. Therefore
I 15% L
A short superstructure or a short deckhouse is so dimensioned that
1< is
A superstructure is a structure that stretches from port to starboard, so that the sides continue the side-plating of the hull.
A deckhouse is a structure placed more or less
symmetrically on the hull, the sides not being in
the same plane as the ships' sides.
A superstructure may be constructed in such a
way, that the top-deck of the superstructure
be-comes the strength-deck of the ship. This will never be so in the case of the top of a deckhouse.
Long superstructures and long deckhouses may both contribute to the longitudinal strength of the
ship.
It is possible that long superstructures and long deckhouses will not contribute to the longitudinal strength of the ship. They are then constructed as "light" structures.
Short superstructures and deckhouses never
con-tribute to the longitudinal strength of the ship.
They are always constructed als "light" structures. The above mentioned structures belong to group I.C. They are by far the most important as regards the strength-calculations and the rules and regu-lations for aluminium-alloy scantlings. Therefore these structures are considered in extenso:
a) Those forming part of the strength-deck' of
Those contributing to the longitudinal strength of the ship. < 7)
Lightly constructed superstructures and
deck-houses. (SS 8. and § 9)
The decision whether the use of aluminiumalloy is economic was treated by Muckle, in his general review of the application of aluminium-alloys to shipbuilding [24], as long'ago as 1943, and to ships' superstructures at a later date [25]. He only
con-sidered the influence of the saving of weight in connection with the cargo carrying capacity and
the saving in power and fuel-costs if the deadweight remained the same. In [2] Muckle based himself on the same cônsiderations, therefore knowingly omit-ting all effects on the strength of the ship. Even so, he took into accOunt the possible reduction in beam due to the lower centre of gravity of the ship. These questions are taken up in the following paragraphs. Not explicitely mentioned in these paragraphs, are the considerations regarding:
The difference in the coefficients of expan-sion between aluminium-alloys and steel, treated by Mucizie in [24] and by Corleti in [26].
The danger of fire and the necessary fire-protection in ships with aluminium-alloy super-structures, a question treated by Venus, Coriett
[27] and Audigé [28].
The difficulties of 'obtaining adequate con-nections between the structural parts of
aluminium-alloy or of aluminium-aluminium-alloy and steel. This question
posed in [1] is treated by a separate panel of
ISC/TC 8/WG -. (See §12, 13 and 14)
The high price of aluminium-alloys.
§ 6. The alumtniwin-alloy top-decks of super'
structures forming part of the strength-deck of the ship
This question may be treated best by making the bold statement, that it has been proved that such a conception is not economically justified. I shall
come back to that point in the next paragraph
where the superstructures and long deckhousescon-tributing to the longitudinal strength of the ship
are described.
The proof of this statement lies in the fact, that Muckle 'in [24] bases himself on calculations for top-decks of deckhouses, wherein the stresses never
or just reach the limit f the allowablestresses. Now
the basis for this calculation is a very rough ap-proximation, and in the discussion of [30], Teif er reproaches Muckle for not taking into consideration the much better modern views of Lyndbn Crawford [31] and H. Bleich [32]'.
But these modern, theories give a view of the
interaction of the superstructure and the hull, which indicates, that as, the bending theory as admitted by Mont go'merie and Muckle is not applicable, the
result is, that the stresses in light-metal
super-structures will be even smaller than those calculated
by Muckle [24]. Therefore' it is safe to assume,
that where increasing the thickness of the super-structure-deckplating has no sense if the
superstruc-ture or deckhouse is only "contributing" to the ships' longitudinal strength (see § 7), it 'certainly has no sense in the supposition that this deckplting should act as strength-deckplating. In that case the economic disadvantages would be still greater.
My observations in the discussion of [30] are therefore still valid. In reality the question is to
find the minimum thickness of the aluminium-alloy topplating on the supposition that the deck-house or superstructure is contributing to the longi-tudinal strength of the ship.
§ 7. Calculation of the thickness of deckplating.
Superstructures or deckhouses contributing to the longitudinal strength of the ship For practical purposes it is possible to base these
calculations on the papers of Montgomerie [29]
and Mucizle [24], [25]. Now the starting point for Mont gomerie was, that the section-modulus of 'the
ship with superstructure should be equal to the
section-modulus of the ship without superstructure.
Both authors indicate furthermofe that they are
consi4ering "light" deckhouses not contributing to the ships' strength.
Nevertheless Muckle in [25] pointed out, thit
Mont gomerie's 'method supposed a "certain" con-tribution to the strength of the ship. The supposition of Mont gomerie that a superstructure-deck takes
up a "certain amount of stress" in relation 'to its distance from the neutral axis, is very disputable, as this deck in his study has no material connection with the hull.
Now even starting from this very unfavourable basis, deckhouses have been constructed to Lloyd's regulations (based upon this study of Mont gomerie)
and were satisfactory. Muckle [25] pointed out,
that when this superstructure or deckhouse-top was constructed 'in aluminium-alloy, the situation be-came even more satisfactory.
The cross-sectional area of the superstructurc
deck "a" is given bij Montgomerie as:
T
a =
jiAo(yo±h)+Io
....
(1)where A0 = cross-sectional area of orIginal struc-ture
1 = moment of inertia of original struc-turc
Yo = distance
from neutral axis of the
original structure to thestrength-deck
h = height of superstructure deck above
main strength-deck
a cross-sectional area of the
superstruc-ture-deck
Muckle proves, that it is possible to derive (by a
similar process of reasoning as that adopted by
Mont gomerie) a formula for the minimum
cross-sectional area of a deck constructed of aluminium-alloy. This formula is:
'
A0{-(y0
+ b) _yo}
2 (2)
Jo
A0(y0+h) +10
where apart from the symbols already mentioned: = stress in the steel structure
= stress in the alloy structure
-E8 = Young's modulus for steel
Ea = Young's modulus for aluminium-alloy Examining this equation (2) it will be seen that under certain circumstances a negative value of "a" may be obtained. This, of course, is not admissible and thus:
(Yo + h)> Yo
Now for Lloyd's quality of aluminium-alloy
= 1.70 (I have calculated 1.65) and - = 3.
7a Ea
Therefore it will be seen that unless h/y0 is equal to or greater than 0,76, the area will be negative. This means that the stress in a superstructure deck, as calculated by Mont gomerie's method, can never rise to the value o, no matter what its cross-sectional area is. It therefore would appear that certainly in a
first tier of superstructures and probably ina second
tier also, the cross-sectional area of the
superstruc-ture-deck is of no importance from the point of
view of longitudinal strength.
Muckle proves [30], that for ships up to 600' in
length, this is true and that in the worst case, the stress in an aluminium house with the same
plate-thickness as steel, is only 74 % of the allowable
stress, while in a steel house the maximum stress rises to 100 % of the allowable stress. In no case does the absolute stress in the aluminium-alloy deck-house reach 55 % of the value of the stress in the steel deckhouse and in larger ships this percentage always remains under 50 %
Now, if all this is true for "light" deckhouses as conceived by Mont gomerie, it certainly will be true
in reality and even more so in the light of the
modern theories of Lyndon Crawford [31] and
Bleich [32].
Muckle has pointed out [24], [25], that this low stress in the aluminium-alloy superstructure-deck, being much lower than the one in a steel
superstruc-ture-deck, must put more load onto the original
strength-deck. This extra load may become so great,
that the buckling limit of the strength-deck may
be passed and he therefore gives a formula, which
'0
indicates what value
must have, so that the
Jobuckling limit of the steel deck is just reached. If
we call = W, we find:
-y
-
3Io + A0 (yo + h)2 + W.h
He then observes, that it is possible to augment the thickness of the steel plates of the strength-deck and thus decrease the thickness of the aluminium-alloy superstructure-deck, as this is economically advisable. The increase in the cross-sectional areaof the steel deck may be:
E0
A0(W.y0-10)a
{W.h+Ao(Yo+h)2+Io} a-, =Io + A.y2 + a..b2
(4)
In giving different values to "a", "a2" may be determined. Finally the aluminium-alloy
super-structure-deck must be safe from buckling, which may be controlled by Muckle's formula:
¡E, o
tO=tSA/_._
(5)where: ta = thickness of deckplating in Al-alloy t0 = thickness of steel deckplating
This control is practically never necessary if the aluminium-alloy deck is covered with wooden deck-planking. As will be pointed out further-on, the
buckling strength of the
aluminium-alloy plates is very much increased in this case. Following Muckle's formulae from 1943, described above, one automatically comes to the conclusions:That with Mont gornerie's theory developed in [29], the determination of the thickness of the aluminium-alloy deckplates will be on the safe side having regard to modern theories about the inter-action between the hull and long superstructures.
That even with the superstructure deckplates
contributing to the longitudinal strength of the
ship, the stresses remain so low, that a reduction of the thickness thus derived may be considered if the steel strength deck is re-inforced at the same time.
That such a procedure will always be followed for economic reasons.
That in that case the steel strength deck will become nearly as thick as the original case where the
presence of a superstructure was not taken into
account.
That indeed, that difference is so small and
the tendency to make the superstructure deck as
thin as possible so great, that it is better to consider
the superstructure in all cases as a "light"
deck-house.
This conclusion is somewhat astonishing, but this is the one, which Montgomerie [29] and Muckle [25] came to and which Muckle took up again in
his publication in 1952 before the Institution of
Naval Architects [30].
Therefore I consider that Tel f er's criticism,
though theoretically soúnd, is unjust and that my
own criticism must be revised in this sense, that it
is perhaps interesting to know how a superstructure
deck contributing to the longitudinal strength of
the ship should be designed, bût that economically speaking, it is better to make a superstructure deck
not contributing to the ships' strength and treat
that problem as Muckle did in [30].
§ 8. The dimensions. of "light" superstructures constructed in aluminium-alloys
The preceding paragraphs have settled thç
ques-tion of the design of superstructures and hâve restricted, thé problem to: How to calculate the
'dimensions of light, long deèkhouses?
.Corlett; [22] has indicated, that the. buckling
stress in an aluminium-álloy deckhouse will be
smaller than in a steel deckhouse' in proportion to
=
At the same time the limit of bucklingEa
will be also be about that of steel X . So,
con-sidering that the buckling of the plates will be
fundamental for this structure, it would be logical
to take equal plate-thicknesses for the steel añd the
aluminium-alloy deckplates.
The stiffness of the whole'ship, however, would be somewhat smaller and therefore, in this case the stresses would become aJittle higher than
Ea
= 0.33.
Coriett calculated that 'the ratio of stress would be 0.35 to 0.37. Therefore hê proposed
ta = 1:10 t8
The limit of buckling however is proportional
to the square of the thickness. Therefore, ta oily
has to be = 6 % higher, so ta = 1.06 t8 iS
sufficient.
In steel construction a surplus of thickness for corrosion has been taken into account. This is not
necessary with aluminium-alloys. Therefore ta t8
might be adopted, basing our calculation on the
buckling limit. (See hereunder).
1Qe have already seen that Wa . 1.65 W8 if W
indicates
.- = section
modulus. If we considerlocal strength, we obtain the following ratios: For equal strength t8 = 1.28 t8.
For equal stiffness t8 = 1.44 t8.
(I83Is,S0tal'3t)
For 50 % moré deflection than steel ta 1.26 t8.
(1. = 2 18, so la = t8)
"The reduction for the fact that an extra
al-lowance for corrosion is not necessary with alumi-nium-alloys, is determined as 'follows:
Light superstructure-deck in steel as, prescribed by Lloyd's (Mont gomerie-basis) for a ship of
L = 180 m is 7.6 mm. The corrosion-allowance herein is estimated to be '1 mm. Plate-thickness withOut corrosion-allowance = 6.6 mm. Therefore
the aluminium-alloy plate-thickness is: For equal strength
1.28 X 6.6 = 8.45mm = 1.11 X 7.6 mm.
For equal stiffness
1.44 X 6.6 = 9.5 mm = 1.24 X 7.6 mm.
For 50 % more deflection
1.26 'X 6.6 = 8.3 mm = 1.09 X 7.6 mm. From the' point of view of local strength the 10 % extra plate-thickness for the aluminium-alloy as proposed by COrI ett seems justified, but the example is given for a big ship, where the combi-nation of local strength and longitudinal strength
means that the load on the aluminium-alloy deck remains appreciably lower than with that combi-nation iû a steel deck. 'With a longitudinal stress of
about' that in a steel deck and Wa '1.65 W8,
we find that the load in the aluminium-alloy deck
in the longitudinal direction will be about
Ï.82 better. For small ships the corrosion-al-lowance is relatively much higher. Therefore it seéms justified to take 't8 = t8, even if the calcu-lation is based on local strength.
For deciding the deckpläte thickness of an alumi-nium-alloy superstructúre-deck, Muckie [30] 'has foll9wed the procedure already discribed in § 7.
Taking into account the smaller stiffness of-the
whole ship, when an aluminium-alloy superstructure is fitted instead of a steel one, he finds that the ratio of buckling-limit to buckling-stress is smaller for the aluminium-alloy than for steel. As' the buckling
limit is' proportional to the square of the
plate-thickness he arrives at the formûla already men-tioned in § 7:
Ea O',
Muckle then calculates 'for different ships'
lengths the necessary plate-thickness for the light-métal superstructure-deck, but it is curious to note thai the 'values given in his table V of [30] do not córrespond with formula (5), which' gives smaller
thicknesses.
-As he does not, take into account the corrosion-allowance, which gives relatively greater plate-'thickneses for small ships, the ratio for sthall
ships becomes greater. than for large ones and the result is, that for small ships, the plate-thickness
of the aluminium-alloy deckplates is too large for two reasons:
The corrosion-allowance for steel need not be added to the plate-thickness for
aluminium-alloys.
The ratio is too large for small ships (see fig. 1 in [30]. Here W1 is the sectional modulus
for a ship with superstructure and W, the
sectional-modulus for a ship withoutsuper-structure.
When correcting the ratio's for as Muckle gives them for corrosion-allowance, then nearly all come to the formula
ta = 1.04 t8.
Muckle then gives some consideration to "elastic buckling". From his experiments and calculations the conclusion may be derived, that the buckling stresses are so small, that with his real ratio's t = = 1.14 t8 and ta = 1.27 t, there will never be any difficulty at all. Taking the plate-thickness of the
aluminium-alloy deck as small at Ea 1.04 to, it is
safe to suppose that even then the buckling limit will not be reached.
Both Corlett [22] and MuckJe [21] take into
account the wood deck-sheathing and its influence on the buckling of the deckplates. As already given
in [1] we find for plates of Y2" thickness and
planked with 2" teak the following ratio's: Unsheathed light metal . . . . 1Unsheathed steel approx. . . 3 J
Sheathed light metal approx. 9.1
Sheated steel approx. 10.4 J
It will be clear therefore, that there is no reason to fear any difficulty in strength for sheathed light-metal decks compared with steel ones, if the un-sheathed decks give the measure of ratio-thickness between aluminium-alloy and steel.
Recapitulating the comparative dimensions to
steel as set up by Corleit and Muckle for light,
long deckhouses, we find these dimensions given in Table I. In the same table the proposed amendments on these figures are given.
The final proposal to ISO/TC 8/WG 1 is not
to take a certain percentage on the thickness of
the steel-deckplating, but to make all unsheathed aluminium-alloy decks OJ mm thicker than steel decks. This proposal is based on the following
reasons:
1) Thin plates receive a relatively greater
ad-ditional thickness than heavy ones. Therefore local strength, which is of more importance with smaller ships, is better allowed for in these ships.
ratio 0.33 ratio 0.87
Based on local streng/h: Equal strength
stiffness
0 % more deflection strength corrected for
cor-rosion allowance
stiffness corrected for
cor-rosion allowance
TABLE I
Requirements for aluminium-alloy
deckhouse-Platin g
Thicknesses are given as the ratio t8/t5, where = thickness of aluminium-alloy plating
Beams and s/i/fences
Section, Eq,l strength = We = 1.61 \01s Cslcu- ,. stiffness
-
la = 3 Islatad on 10 % deflection
-
¡a = 2 IsEqeal buckling
-
¡a = LI IsI) The lower numbers ore for L = ISO m, the higher ones for L = 60
es. 2) This is true for both types of ships (I tier and 2 tiar, of beams).
The addition of 0.5 mm thickness, corre-sponds in ships varying from 60
m to 180 m in
length to an additional percentage in longitudinalstrength of 8 % to 6 %. These values are bigger than the 4 % allowed in the criticism of Muckle's
values.
As already mentioned the combination of
local strength and longitudinal strength of the ship gives an advantage to the aluminium-alloy struc-ture. Therefore the additional longitudinal strength percentage of 8 % to 6 %, is certainly covered by the 11 % local-strength addition as corrected for
corrosion-allowance.
For sheathed decks, the proposal is to take for aluminium-alloy decks the same thickness as for
steel decks. As MuckJe has taken no account of wood sheathing, it is safe to assume that the 4 % greater thickness as asked for by his method will be largely counterbalanced by the better strength-ratio (0.87 against 0.33) of the sheated decks in alumi-nium-alloy.
Concerning the sides of the deckhouses M lick! e
[25] proposes t = 1.15 t8 and GoncEE [22] ta to.
Both authors are illogical, but it may be that Conlef t takes into account the less important local loads and therefore drops his additional 10 %.
Va = 1.61 Ws; with un-sheathed deck, and heavy beams Va = 2 Vs. = ,, steel plating Item Corlett ta/is = Muckle io/is =
Based oie buckling limit:
Unsheathed deckplating 1.10 1.06 to 1.25')
Taking The longitudinal bending into ' The corrosion allowance account I. Both factors
1.06
Working with Muckle's J 1 Tier 1.07 to 1.241) formula I 2 Tiers. 1.14 to 1.271)
Taking into account "elastic buckling'
Sheathed deckplating
corrected for
cor-rosion allowance 1.00 1.00 1.12 to 1.21) 1.04 As criticized ia/is = 1.00 1.00 1.042) l.04) 1.00 1.00 1.00 1.28 1.44 1.26 1.24
Logically one can say, that the load on the
vertic-al vertic-aluminium-vertic-alloy house side, will be about 50 %
of that in the case: of a steel house side. As the
limits of buckling of the aluminium-alloy plating and steel plating are in the proportion of Ea : Ea =
= 1: 3, we may corsider that la = 0.5 X 3 18
=
= 1.5 L and therefore ta = 1.14 t8. This ratio cor-rected for corrosion-allowance on steel only, gives as a result:
ta 1.0 t8.
The proposal is to do the -same as in the case of
the deckplating and for the same reasons adopt an additional thickness of plating in aluminium-alloy of 0.5 mm.
The deckbeams must be considered- separately: Muc/de [25] gives for beams of:
a. Equalstrength Wa 1.65 (1.70) W8
- b. Equal stiffness 1 3 18
-c. 50%more deflection f
-
2 LCorlett [22] bases his tables on deflections of the beam of 1/500 with 2 feet cargo height and a
cargo density of 50 cubft/ton. Based on steel
scantlings as prescribed by Lloyd's, Corlett finds:TABLE II
means (W
=)
= sectional modulus ofplate + profile of the aluminium-alloy strUcture.
W5pof means (W = ) = sectional modulus
of profile of the steel structure.
We -see that Corlett allows deflections in the light-alloy beams more than twice as great as in
the steel- ones. = constant 1.7.
Bureau Ventas prescribes only that - = 1.5.
This gives still greater deflections of the beams as
compared with those of Cori ett, as
-'a
1.35.
It is difficult to make a choice from these three
propositions. The fact must be considered, that
with more than one tier of light-metal superstruc-ture decks, one of these decks is submittted to, what
Muckie [30] calls, the "eñd-effect" of the
deck-house and which -phenomenon is very clearly
described by Van der Neut [33]. The beams not
only must be strong enough, but also form a certain solid ground fOr the big "end-effect" forces of the
deckhouse above.
-A superposition- of the deflections of several
decks, where more tiers of decks are present, may
result in inadmissible deflections of the highest
deckhouse. Therefthe the stiffness should not be töo small and at least Wag8
= 165
1V,88 should beobtained. .
For heavy bulb-sections the progress of I and W is equal. For wood deck-sheathing the deflection is
diminished. If there is no planking there should
be a tendency in big ships having deck-beams in
thé
f rm of bulb-sections to obtain minithum
deflections -and in this case lVat0, = 2 IVSO8 should
be proposed.
Muckle [30] studies the stiffeners of- the -sides
and around the openings for windows in
deck-houses. Now, de to the end-effect the forces taken by the front bulkheads and for a distance of about the breadth of the deckhouse along the longitudinal wall, will be between 70 % to 90 % of the total
vertical forces- working on the deckhouse-sides. In these parts the stiffeners have an important duty -to fulfil. If -there are "openings" in -the side, the normally calculated stiffeners should have an
in-crease in cross-sectional area of about 2.0-% to 30 %
according - to Muc/zie. Neither Coriett nor Bureau Ventas give indications on this subject.
Conlett gives for the normal stif-fener the
for-mula Wa = 1.-65 W8. The table given in [22] for the dimensions of these stiffeners has everywhere W8 > 1.65 W8 and la varying from 1.75 to 2.1 1. Muc/zle [30] indicates that stiffeners must be
dimensioned on the total stiffness of plate + section, which is interpreted by Robinson [34], as Ia = 2 1.
This should of course be l = 3 L for the section
only. Muckle's requiremçnts seem too heavy in this case, the more so as the -weight of the aluminium-alloy deckhoúse is much less than that of the steel deckhouse. Furthermore there is no reason, why we should hold to the requirement of equal stiffness as a little greater. deflection of the side will do no harm. Based on the buckling limit with a weight
load of 50 % of the steel house, one finds Ia = 0.5
X 3 Is 1.5 Is. Therefore it is proposed to accept here also the formula W = 1.65 W8 (see Corlett's
proposal). -
-The bulkheads inside the deckhouses and their stiffeners must remain relatively heavy. The fittings
of the house which are
not influenced by the
material in which the house is constructed, will remain of the same weight and that load will be
taken, for the most part, by the interior bulkheads. The corrosion-factor also plays a lesser part in considerations concerning the dimensions of these
-Il
Free length of beam: 1: 10' 20'
W'88,/ 1.80 1.94
1.70 1.70
1V0,.O1/ T"prof L63 1.90
inside structures. Therefore instead of the proposal
= t8 + 0.5 mm as for the unsheathed
deck-plating and the outer walls, the proposal here
should beta = t8 + 1mm.
For the stiffeners of these inner bulkheads the
same considerations as
for the other stiffeners
remain valid and therefore Wa = 1.65 W8.
For pillars, which are subjected only to
column-buckling it is clear that la = 318. Robinson [34]
gives a general survey of the comparative
dimen-NOTES: All scantlings for aluminium-alloys are based on the
relevant dimensions of steel. Thus I and i and W8 are the thickness, the moment of inertia and the sectional modulus for a steel structure.
COMMENTS ON THE TABLE ARE:
The higher figures for the shorter deckhouses.
(S) is for short deckhouses less than 0.20 L. (L) is for long deckhouse, greater than 0.20 L.
a) The lower W5 is permitted with a % Mg-alloy in light-hardened
condition or Al - Mg - Si heat treated sections.
TABLE III
Requirements fore aluminium-alloy deckhouses
sions of light, long deckhouses in his paper read at Naples. He therein also compares the regulations laid down by the different classification societies. The table VIII of his paper is completed here with the proposals as put forward to ISO/TC 8/WG 1
and the result is given in toto as table 1H of this
study.
This proposal is simple and easy to handle. It is on the safe side as the breaking point of the
alumi-nium-alloy is the minimum one prescribed by
Lloyds, whereas for steel the meanbreaking point is taken.
3.0 I here applied to columns undergoing compression stresses stresses such as pillars etc.
For structural units under local loading and where equal
de-flections are required.
O) If the deck plating is connected to wood sheathing at centre of
every beamspace, the thickness may be the same as for steel. 2W8 here applied to unsheathed decks of large ships, where the
deck-beams are bulb-sections.
3.0 1, here applied to columns undergoing compression stresses (Pillars).
Item Corlert Muckle
Norske Ventas 2) Bureau Ventas Registro Italiano In ac-cordance with Lloyd's practsce Proposal to ISO/TC 8 W.G. i Deckhouse-sides Plating 1.10 t8 1.15 t5
j
' 1.20 t4 (S) 1.35 18 (L) 1.20 t8 J 1.45 t8i
1.25 . 1.15 t8 t8 + 0.5 mm Stiffeners 1.70 W3 11.30 W8 (S) . 1.50 W8 (L) ¡1.30 W83) 11.20 W5 2.25 W8 1.70 W8 1.65 Ws Deckhouse-fronts 1.20 18 1.50 18 1.20 t8 18 + 0.5 mm Plating 1.10 Stiffeners . 2.0 W. { 1:30 W8 (S) isø W8 (L) 2.25 w8 2.00 W. 1.65 W3 Dec /zplating Sheathed 18 1.12-1.25 t8 1.20/3 (L) 1.20 t8 1.10 t8 O) t8 Unsheathed .. . . 1.10 t8 1.12-1.25 8 1 1.20 t4 (S) 1,35 t8 (L) 1.35 18 1.15 t8 t8 -+- 0.5 mm 1) Inner-bulkheads Lined Unlined t8+0.SmlTit8 + 1.0mm Stiffeners 2.0 18 2.0 I 2.0 18 1.65 W,5 Deckbeams . . . . 1.70 Ws ¡1.30 W8 (S) . 1.50 W8 (L) 1.50 W3 2.25 W8 1.70 W. 1.65 à 2 W. 7)Beams and girders 2.0 '8 2.0 18 3.013 3.0134)
3.0I)
2.0 si
1 1.65 W,83.0 188)Short superstructures and short deckhouses
As the
corrosion-allowance for thin plates isrelatively more important than for heavy plates,
the application of the above-mentioned rules for long deckhouses, to short superstructures and deck-houses will give the latter a greater factor of safety. As, on the other hand, the local loadings of these deckhouses are of the same order as those for long deckhouses, it is considered prudent, not to suggest a decrease in the thickness of the plating of these short superstructures and deckhouses, relative to the
long ones.
Therefore t = t, + 0.5 mm.
For the stiffeners W = 1.65 W8 is proposed.
The question of when a deckhouse is short or
long depends on the prescriptions of the classifi-cation societies. Norske Verjus considers a
deck-house short when I < 0.2 L, Lloyds when i <
0.15 L and Lyndon Crawford when I < 0.35 L
[31].
Corrosion-resisting coverings and linings of insulated cargo-spaces, tanks, kitchens, etc. and the fabrication of ships' fittings exposed
to corrosion
These types of application are indicated in the groups II. E and II. F. As the corrosion-resisting
qualities of the aluminium-alloys are the reason
why this material is chosen, there will be no need
to discuss the desirability of using it. Under this heading also falls the use of aluminium-foils for
insulating purposes.
The advantages of aluminium-alloys in this field
are more e'specially:
Gain in hold capacity and weight. Longevity of the coverings. Easy repair.
Gain in fire-protection as regards cork. Hygienic.
Resistance against corrosion by oil or wine. The thickness of the plating varies in this case
from 2 to 4.5 mm, depending on the supports
available. Ceilings
are constructed in plates of
1.5 mm to 2 mm. The bottomplating nearly always
has an increased thickness of 1.5 mm above the
wall-plating. The bottoms of cargo-spaces are still
Suggestions about the riveting between steel and aluminium alloy ship's structures.
§ 12. Introduction of this special subject
The only possible means of connect ng steel to aluminium-alloy structures are riveting and bolting.
In view of the difficulties arising from riveting
thick plates together, it might, first of all, be
sug-gested, that the idea of bolting together ships'
structures should not be disregarded without careful
study of the subject. Indeed fitted-bolts may be
cadmium plated or protected against corrosion in
often made in wood, as the wear and tear is great. For refrigerated cargo-ships see [17], for fishing
vessels [14], [15] and [16]. Good guidance for
the upkeep of these installations is to be found in [14] and [35]. The connection of these thin sheets by means of "cadmized" screws etc. is outside the scope of this study. Information about port-holes and windows is given in [18]. An increase of the
corrosion-resistance is
often brought about by
anodizing or other chemical treatment. Bimetallic contacts must be prevented at all costs.
§ Il.
Special ships' fittings, which must be lightand resistant against corrosion
These fittings are listed in the groups III. G and III. H. The use of aluminium-alloys may be stimu-lated by the ease of handling during the fitting of these parts. This especially is the case with
venti-lator-ducts, pipe-lines etc. The transport of very
corrosive fluids is best done by light-metal pipes. The dimensions are determined by normal
strength-calculations. Welding of pipes instead of flanging them and bolting them afterwards is in-creasingly used.
The use of light-metal lifeboats on board large passenger-ships saves an enormous amount of
top-weight, and the boats' davits may also be
con-structed in aluminium-alloy. Light-metal ships'
funnels have the same advantages Resistance against
corrosion at high temperatures is a question that
still needs some experimental research. Up to the present the full corrosion-allowance on the steel
has been kept for the light-metal funnels. The
upkeep of these big light-alloy funnels is described
in [19] and [20].
The scantlings of lifeboats and funnels may be determined by direct strength calculation. Though welding gives a saving in the quantity of material used, thin plates are more easily riveted and thus the above mentioned gain is small. Here the normal practice is to make the rivet spacing one diameter less than in the corresponding steel-riveted connec-tionS [36].
The choice of the aluminium-alloys used on
board is given in [1] and forms a part of the study of a special panel of ISO/TC 8/WG 1.
such a way, that they can take the place of rivets, even for great plate thicknesses.
But the object of this study is, to give informa-tion about the riveting of connecinforma-tions between steel and aluminium-alloys. Points of interest are:
Choice of rivet (see § 13).
Precautions against corrosion (see § 13 en 14). Execution of riveted construction (see § 14).
TABLE IV
General Survey of the possibilities, advantages and disadvantages of the use of hot- or cold riveted connections betiveen steel and
aluminium -alloy ships' structures
Riveting
Rivets
Rivet material
compared to plate material corrosion
Effect of
shrinkage of
rivet
Effect of tempe- rature of rivet on
.
material and packing
Max.
.
of rivet
.
Effect of rivet head
. . Pitch of rivets Corrosion of joint Hot riveted Steel rivets
If not alumited chance of much galvanic cor- rosion
Good shrinkage strong clamping of plates. Rivet little elongation and niuch heat con- duction. Tight joints, Temperature too high for plates, too iow for rivets,
Destruction of packing by heat. No alumiting possible. temp.: ± 6000 C. still too low Above 6 mm
Snaphead on the steel plate-side. With galvanised washer on Al-plate side
Normal
As packing is destroy- ed corrosion nearly in- evitable. Heat durable packing.
Al-alloy rivets
Rivet metal as far as possible the same as plate material to avoid galvanic corrosion
Good shrinkage. Good to doubtful clamping of plates. Rivet large elongation and much heat conduction. Tight joints, but doubtful with heavy plates. Dif- ficult to get the right temperature.
Destruction of packing by heat. Alumiting of steel plate nearly impos- sible. temp.: ± 5000 C. Above 8 mm
Snaphead on the Al-plate side. With galvanised washer on
Al..
plate side
About i d. smaller than with steel rivets, More rows of rivets
As packing is destroy- ed corrosion nearly in- evitable, Heat durable packing, not flexible enough.
Cold riveted
Steel rivets
If steel surface> than Al-surface: Snaphead on the side of steel plate if corrosion on that side. Corrosion smaller than with Al-rivets, Small corrosion on Al- plates.
No shrinkage; strong rivet with small 0 Certain strength in direction of rivet-axis, good if corrosion starts between plates.
Alumiting, cadmium plating and galvanizing possible
8 mm excep- tionally 10 mm
Snapheaci on steel plate side
Normal. More rows than in steel construction due to small diameters Glue packing and paste without packing is pos- sible. Packing must be flexible and not hygro- scopie. Chemical and electrical isolating.
Al-alloy rivets
1f Al-surface > steel surface: Snaphead on the side of Al-plate if corrosion on that side. Bigger corrosion than with steel rivets. Big corrosion of rivet duc to steel plate. No shrinkage. Weak rivet with big 0 Doubtful strength and too much elongation in axial-direction. If cor- rosion starts no rigi- clity.
Alumiting, cadmium plating and galvanizing is possible
16 mm excep- tionally 22 mm Snaphead on Al-plate side
About i d. smaller than with steel rivets, More rows of rivets Glue packing and paste without packing is pos- sible. Packing must be flexible and not hygro- scopie.
Chemical and
§ 13. Choice of the riveted-connection
(see Bibliography [37]-[44])
We have the following choices for riveting the
connections of steel to aluminium-alloy ships'
structures.
Use of hot-riveted steel rivets. Use of hot-riveted AL-rivets. Use of cold-riveted steel rivets. Use of cold-riveted A1.-rivets.
Table IV gives a general view of the possibilities; advantages and disadvantages of the use of these different methods; also the corrosion aspect.
§ 14. Execution of the steel-riveted joint
The use of steel rivets for this kind of connection is to be generally recommended (see Table IV). As
far as possible these steel rivets should be cold
riveted. If hot-riveting is unavoidable, the riveting
temperature must be kept as low as possible for
executing the riveting. In all cases these temper-atures should not be more than 6000 C.
If such high temperatures are necessary, galva-nised washers should be used.
H. E. Jaeger: 'Practical possibilities of constructional
appli-cations of aluminium-alloys to ship construction". Report
No 3 of the Netherlands' Research Centre T.N.0. for
Shipbuilding and Navigation, March 1951. See also Schip
en Werf", 27th April and 11th May 1951.
W. Muckje: 'The influence of weight-saving on dimensions
and form". The Shipbuilder, July 1952.
A. Pyszka: "Vedetten aus Leichtmetall", Zeitschrift der
Alu-minium Zentrale 1953, pag. 37.
'Aluminium Yacht of welded construction", Weld- and Metal-fabricatión, August 1953.
E. Helsengreen: 'Eine Vergleichsrechnung Stahl-Aluminium für
ein schnelles Revierschiff", Schiff und Hafen, July 1954. 'Leichtmetallmerkel", Hansa 1951, pag. 324.
H. G. Auerswaldand A. Scyrnans/ci: 'Schiebebalken aus Leicht-metall", Schiff und Hafen, September 1953.
W. Surmaier: Möglichkeiten des Leichtmetalleinsatzes im
Bin-nenschiffbau", Zeitschrift für Binnenschiffahrt, February
1953.
K. Suppus: "Die Entwicklung der Leichtmetall
Lukenabdeckun-gen", Zeitschrift für Binnenschiffahrt, February 1953.
Tb. Domes: "Spezielle rechnerische Grundlagen für die
Leicht-metallanwendung im Schiffbau", Jahrbuch der Schiffbau-technischen Gesellschaft 1952.
K. Su/ter: "Erfahrungen im Berechnen von
LeichtmetaÍlkon-struktionen", Wirtschaft und Technik im Transport,
July-November 1951.
"Rettungsbote",Hansa 1951, pag. 1672.
W. Fiedler: ,,Leichtmetallkonstruktionen für den Schiffbau",
Hansa 1950, pag. 1484.
F. Kraus and A. Müller-Busse: "Der Leichtmetallfischraum der
Arktis", Hansa 1951, pag. 61 2
Bibliography
Cold-riveted
rivets may with
advantage bealumited, galvanised or cadmium plated. The rivet hole must be filled up with paste or paint. Rivet-holes must be made with the utmost care and the rivet-fitting must be as perfect as possible.
All packings must be flexible. Thereforé paste without fibre may be recommended as far as pos-sible. The more so, as every fibrous canvas is apt to be hygroscopic, so helping to introdiice corrosion. All paste should clear the joint for 2 or 3/mm.
The snaphead should be at the steel-plate sid'e of the joint. This side should be the most corrosive sided The joint, must be well above places where seawater may accumulate. The distance from the
bottom of the joint to a wood deck should be at
least 50 mm.
For hot-riveted joints the breadth of the overlap may be made somewhat larger than necessary from the strength point of view. In all cases > 75 mm.
Caulking of these "mixed" structures is not pos-sible. For obtaining watertightness, good "plane"
surfaces of both the steel and aluminium-alloy
plate. are necessary. After riveting, varnishing of
the joint with lacquers etc. may help to achieve
watertightness.
W. Fiedler: "Leichtmetallfischräume", Hansa 1951, pag. 1673.
"Der Arktis nach zweijähriger Betriebszeit", Hansa 1952,
pag. 428.
P. VidaI: "L'allègement de l'isolation è bord des navires
frigori-fiques", Bulletin Technique du Bureau Ventas, January 1951.
"Schiffsfenster" DIN. 81601, April 1953.
A. Cherrier: "Les cheminées de navires en alliage léger", Revue
de l'Aluminium, January 1952.
"Lightmetal ships' funnels", Shipping World 1953, page 186.
'W. -Muc/de: 'Resistance to buckling of light-alloy plates",
Trans. N.E.C., 5 March 1948.
E. C. B. Corle/E: "Aluminium as a shipbuilding material",
Trans. N.E.C., 22 February 1952.
F. S. M. Brierly and J. E. Tom!inson: "Some recent develop-ments in the welding of aluminium-alloys and their future applications in the shipbuilding industry", British Welding Journal, 1954.
W. Muckle: "Some considerations on the application of light
alloys to -ship-construction", Trans. N.E.C., 20 December
1943.
'W. Muckle: "Application of light-alloys to superstructures of
ships". Trans. N.E.C., 26 April 1946.
E. C. B. Corle/E: "Thermal expansion effects in composite ships",
1.N.A. 1950.
J. Venus and E. C. B. Corle/E: "Fire protection in passenger ships", I.N.A. 1953.
A.Audigé: "Etudes récentes sur la protection cntre l'incendie la prévention dans la construction", A.T.M.A. 1954.'
J. Montgomerie: "The scantlings of light superstructures",
I.N-.A. 1915.
W.Muc'kle: "The scantlings of long deckhouses constructed of aluminium alloy", I.N.A. 1952.
Lyndon Crawford: "Theory of long ships' superstructures", 38. S.N.A.M.E. 1950.
H. B1eic: "A study on the strùctural acioh of superstructures on ships Ship Structure Committee No SSC 48 Washing ton,, January 1953.
A. - van der Nevé: "Beschouwingen over vormgeving van gelaste scheepsconstructies", Lassymposium der N.V.L., Utrecht, 26 October 1951.
L. M. C. Robinson: "Aluminium in ships' structures; .a review
of current practice". Trans. of the International Maritime
Congress, Naples, 30 September 1954.
-R. Klingholz: "Neuartiges Montagevexfahren an Bord von Schif-fen", Hansa 1952, page 1590.
J. E. Temple: "Handbook of structural designs in aluminium-alloys", Editiòñ: James Booth & Co.
37.- Aluminium Zentrale Düsseldorf, Aluminium Mérkblatt V 5.
J Reiprich: "Werkstoff- und Betriebstechnische Grundlagen für die Aluminiuthverwendung aúf Schiffen", Alumiñium im Schiffbau 19,53, Aluminium Zentrale, Düsseldorf. 39. "L'ùltilisation de l'Aluminium et de ses alliages en -construction.
navale". L'aluminium français, February 1954.
40 C V Boy/cia and M L Sellers Practical problems relative to the use of aluminium-alloys in ship construction'. S.N.A.M.E., New York 1953.
"Aluminium for marine uses", Ships' structures of the British Auninium Co. Ltd. 1953g
A.D.A. Information Bulletin No. 20, December 1952. W. Sandow: "La protection contre la corrosion des constructions
en métal léger", Aluminium Suisse; page 198, November 1954.
J. M. Whiteford: Aluminium' and bimetallic- joints", Shipping World, 2 September 1953.