SOME MODEL AND FULL SCALE
EXPERIMENTS ON SIDE LAUNCHING
B
F. H. TODD, B.Sc., Ph.D., Associate Member
& E. LAWS
A Paper read before the North East Coast Institution of Engineers and Shipbuilders in Newcastle upon Tyne
on the 28th April, 1947, with the discuasion and
correspondence upon it, and the Authors' reply thereto.
(Excerpt from the Institution Transactions Vol. 63.)
NEWCASTLE UPON TYNI
PUBLISflED BY ThE NORTH EAST COAS1 UT1ON
OF ENGINEERS AND SHIPBUILDERS, BOL»EG 1tLL
LONDON
E. & F. N. SPON LIMITED, 57, HAYMARKET, SW.1
THE INSTITUTION IS NOT RESPONSIBLE FOR THE STATEMENTS MADE, NOR FOR THE OPINIONS EXPRESSED, IN THIS PAPER, DISCUSSION AND AUTHORS' REPLY.
ERRATA
p. 346. Table 1, fourth column For "in." read "ft." p. 353. Second paragraph:
For "Table i " read "Table 2"
p. 358. Seventh paragraph: For "cage" read "case"MADE AND PRINTED IN GREAT BRITAIN Printed by Charles Birchall & Sons. Ltd., Liverpool 2.
2W
SOME MODEL AND FULL SCALE
EXPERIMENTS ON
SIDE
LAUNCHING
By F. H. TODD,* B.Sc., Ph.D., Associate Member,
and E. LAWS.
28th April, 1947
SYNoPsIs.The building of the Mulberry Harbour involved many extremely difficult technical and constructional problems. Not least among the former was
that of transferring some of the large concrete units to the water. These had
to be built at a number of places around the coast, wherever sites, labour and
materials were available. Some of them were built in dry docks and subsequently floated out, but others had to be constructed on land and, because of restricted water space, launched sideways. Among the large amount of experiment work carried out in the Ship Division of N.P.L. on the whole Mulberry Scheme, were some model launching experiments. These are described in this paper, together with observations taken during the launches of the actual units. The agreement is such as to give confidence in the use of models for the investigation of anyfuture launching problems.
1.Methods of Side Launching.
SIDE
launching is usually, although not always, adopted in place of end launching because of lack of water space. The problems raised
by the two methods are quite different.
In end launching the ship
usually enters the water at considerable speed, since the run is long and the water resistance relatively small. Although there must be sufficient water space, it is usually not unlimited, and one of the problems is to arrest the
motion of the ship, which is commonly done by drags on anchors. The
chief danger in such a launch is the possibility of structural damage due to tipping about the way ends.
In side launching, the run is normally short, and the high water resistance experienced by the ship when in the water and movingsideways rapidly stops the motion, so that no drags or anchors are as arule necessary. The chief dangers which now arise are of structural damage due to the drop from the quay edge, from shipping water on the outward roll, or from the side of the ship hitting the quay on the reverse roll.
Four methods of side launching have been used, and are illustrated in
Fig.
1:-(a)Fixed ways (Fig. 1 (a))
The standing ways reach just to the quay edge, and are fixed right up to their ends. As the ship moves down the ways, she finally pivots about the way ends and drops off into the water, at aconsiderable angle
to the verticalperhaps 45 degrees. This method is commonly used
in this country for launching barges, tugs and trawlers into canals
or narrow rivers.
(b) Tilting Ways (Fig. 1 (b))
The last length of the standing ways is allowed to tilt, so that the trans-ference of the ship to the water is eased somewhat. This method is fairly commonly used in America for Great Lakes vessels and other
craft.
*Princjpal Scientific Officer, Ship Division, 1P.L. fExperimental Officer, Ship Division, N.P.L.
TARLE i o, Ship Length Displace-Grease Run to Speed at ment Pressure Dock Dock Coefficient. GM Edge Edge I Friction Tons Tons. Sq. in. Feet F.P.S. Declivity I jus, per Ft.
Drop off Inc/ies
Heel Derecs r' z o 'n Submarine 300' 938 178 32 0 04 I 27 1slu 3 48 t-, n Cnrgo .
-2,465 261 50 0031 1f -13 n n Cargo . 2,565 272 50 1+ 'n 30 ut z Minesweeper 351 225 28'6'-.
n 14 13 -z -t (n o Destroyer-.
1,330 2'59 46'6' -0020 14 z 3 (M on 80' 740 175 . 189 13f 0029 37 2 * Nil Z-- z Concrete C) z Pontoon 202' 3,155234
¡ 91 19f O'03 2 '8 1f Nil 12 z nSOME - MODEL AND FULL SCALE EXI'ERIMENTS ON SIDE LAUNCHING 347
Building on piles (Fig. 1 (c)
Certain concrete barges were built alongside a quay on concrete piles resting on the river bed. When a barge was complete, hand jacks were used to move her slowly outwards. The piles finally overbalanced
and the barge dropped vertically into the water with no further outward
movement.
Fixed ways built right out under the water (Fig. i (d))
This method has not been very much used in the past, but was adopted in both the cases described in this paper. Under this condition the vessel moving down the ways experiences a very high resistance when she reaches the water. The problem is rather to keep her moving long enough to float off the ways than of arresting her motion. If this can be done successfully, it obviates any risk of damage during the launch, and this consideration was important in the case of these large concrete structures.
2.Sorne Statistics of Side Launches
A number of papers have been read in the U.S.A. on the subject of side launching some of the more important statistics are given in Table 1.
3.The Origin of the Conway Tests
The problem of launching the first of these craft was brought to the notice
of the staff of Ship Division in January, 1943, by Mr. H. Torys Hughes, acting on behalf of the Director of Transportation (Tn. 5.), War Office.
A large concrete pontoon, some 200 feet in length, was then being built at Conway, North Wales, as a unit of experimental harbour equipment, an idea which was in fact to grow into the famous Mulberry Harbour. It had to be
launched sideways into a narrow river, in which a rapid current flowed between high and low water. The unit would have a launching weight of about 3,155 tons, and the width of the river being only about 400 feet, the authorities were very anxious to explore the possibilities of damage to houses on the opposite shore, in order that any necessary action might be taken to protect or insure them. Ship Division was, therefore, asked to construct a model of the river and carry out launching tests to observe the wave formation on the far shore.
The only launching experiments previously made at N.P.L. were on a lifeboat launched in the usual endways fashion. We were, therefore, rather doubtful as to the application of the results of the model tests to the full size event, and it was agreed with Mr. Hughes that if we carried out the model
work, he would give us every facility to obtain the equivalent data for the actual ship. It is necessary to point out here that we were in no way responsible for any of the actual pontoon launching arrangements, the model experiments being solely directed towards predicting the wave formation and its possible effects on the opposite shore. The arrangements for launching and all the
necessary calculations were made by the contractors, Messrs. Holloway Bros.
(London) Ltd., under the general direction of Mr. Torys Hughes. 4.Arrangements for Ship launch
The pontoon was a concrete structure of the following principal
dimen-sions
:-Length 202' 7"
Beam 45' 0
Depth 24' 0"
Draught 14' 4)
Launching weight 3,155 tons
348 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCSIING
The cross sections were everywhere rectangular, and in plan form there was
a long parallel centre part, covering 68 per cent. of the length, rounded in quite sharply at each end. The photograph reproduced in Fig. 2 gives a general idea of the whole unit with its large steel girder superstructure.
The lay-out of the standing ways and the general arrangement of the site are shown in Fig. 3. There were ten ways in all, each 3' 0" wide, giving an average pressure of 2 34 tons per square foot, the ways being cambered to a constant radius of 1,480 feet, with a slope at the centre line of the ship before the launch of i in 16, and at the extreme end of the ways i in 5, the mean declivity being i in loi.
The ways were greased with Tallene, launching lubricant. Sperm oil was
added on top before the actual launch.
5.Model Experiments
A model of the pontoon, N.P.L. No. 2207, was made in paraffin wax to a scale of 1/20th full size, being ballasted with weights to give the correct scale displacement of 861 lb. The weights were so distributed that the pontoon floated level, on even keel, and had practically the correct GM to scale as determined by inclining experiments. Some rolling experiments were also made in an attempt to ensure that the model had the same transverse radius of gyration as the pontoon. The shortest period which could be obtained on the model was equivalent to 12 5 seconds for the pontoon, as compared with the estimated period of II 25 seconds, and in view of the latter's admittedly approximate character, this was accepted.
Before the experiments were begun, calculations had been made of the
probable speed of entry of the pontoon into the water. These were based upon
a coefficient of friction for the grease of O 025, given by Mr. Hughes as the result of experiments using approximately the same pressure per square foot between the surfaces as on the pontoon, namely 2 34 tons per square foot. These calculations showed that the vessel would probably touch water at a
speed of 18 feet per second, would continue to accelerate for a very short time
until a speed of about 2lft. /second was reached, arid that the water resistance
would then cause a fairly rapid deceleration. The remaining work was carried
out on the assumption that the speed of entry would be about 18 feet/second. It subsequently transpired that the experiments to determine the coefficient of friction had been made with Russian tallow, whereas the launch was to be
made with Tallene, launching lubricant and sperm oil, and it was then decided
to extend the model experiments to cover a range of speeds above and below 18 feet per second.
A considerable amount of preliminary work was done on the launching ways, using different areas of surface and types of grease. It will be obvious that if an exact sca]e model were made of the ways, the pressure per square foot would only be 1/20th of that on the ship. In the first place, therefore, experiments were made with only two standing ways, the sliding ways being of just the correct length to give the same absolute pressure on the grease
surface. All these experiments were failures, and it was finally found necessary
to use roller bearings.
The final arrangement of the model apparatus is shown in Figs. 4 and 5. The standing ways were cambered, having a constant radius for the actual
model ways of 74 feet, and consisted of two ways, each 1 inches wide spaced equally 3 25 feet both sides of the middle transverse plane of the model. The ways were of wood, made and supplied by Messrs. Holloway Brothers. The
SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING 349
The sliding ways consisted of two iron bars, 1 inch square, fixed to the
under-side of the model, and inset into each bar were tworoller bearings, projecting
about ," below the bar and having a flange width of inch. The sliding ways
were so shaped and fitted that the model was correctly placed as regards height above ways and attitude. In addition, ten dummy sliding ways were
attached to the underside of the model in order that the resistance of the
launching ways might he comparable in ship and model.
In the remaining description of model tests throughout this paper, all dimensions, speeds and times are given in terms of ship figures, except where
specifically stated to the contrary, as being then more intelligible and more
easily comparable with the full-sized observations.
The experiments were carried out in the shallow end of the New Tank, the water level in the first instance being arranged to correspond to a height of tide above sitc datum of 26 feet. The space between the standing ways was boarded in to very nearly the level of the grease surface for the pontoon, this being the approximate level of the ground and sand at the launching site at that particular time. The shape of the river bed was reproduced
approxi-mately by boarding laid on steel scaffolding, and the beach, roadway, garden
wall and houses on the opposite bank were made correct to scale for a short length immediately opposite the centre of the launch ways. The general lay-out for the experiments can be seen fromFig. 5.
It was found that if the model were simply released and allowed to move down
the ways under gravity only, the speed at entry corresponded to about 206 feet per second, or rather more than the probable speed of the pontoon. To obtain the lower speeds, weights were dragged behind the model during the first part of the launch and the drag cord released by a special automatic trip just before the model entered the water, after which she was quite free from
any restraint. To reach higher speeds, a wire was taken around the aft side
of the launching cradle, brought forward diagonally to points nearly at the water level at each side and led over swivelling pulleys up to two vertical pulleys, the two ends carrying scale pans on which weights could he placed.
When the model was released the tension in the wire gave it additional accelera-tion, and just before it entered the water the wire became taut transversely across
the ways and the model ran clear.
A series of tests was made with different drag and acceleration weights to cover the probable ship speeds of entry. During each experiment records were taken of the distance travelled and the time. Observations were made of the time from release to the model entering the water, and passing the way ends, and to the arrival of the first wave on the opposite shore. Cinc films
were taken of the launch, and also of the waves on the beach opposite, a grid being erected at each point for future analysis.
6.Results of Experime?!ts with 26 ft. tide
The speeds of entry varied from 17 to 25 feet per second, and the model ran well clear of the ways on every occasion. As she left the way ends she tipped outwards to an angle of about 12 degrees, and the freeboard left from the wave surface at the hull to the deck was only about 3' 6". This stressed the necessity of having all openings in the deck made temporarily watertight for the launch. The waves travelled across the river to reach the far shore at times varying between 33 and 42 seconds after the release of the ship, according to the speed of entry into the water. The first wave had a height
of 3' 6" to 4' 0" at the shore ; it ran up the beach, across the roadway and broke
against the low garden wall. Some water spilled over into the gardens, but had no force behind it after reaching the wall. The " beach" in the
350 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
the wave both because of its roughness and by actual absorption of water
through the shingle. To quote from the first report of 20th March, 1943
"Nothing in the experiments indicated thatany danger was to be feared to the ship herself during the launch from any hydrodynamic forces . . . nor does
any damage to the opposite shore seem likely beyond surface damage to plants,
etc., in the gardens . . . It would be a reasonable precaution to block up
all gateways and openings in the garden wall so as to limit the water entering
the gardens to any which might splash over the walls." How well these
predictions were borne out will be evident when we consider the ship launch.
7.Results of Experiments with 24 ft. tide
The original date proposed for the launch couldnot be kept, and on the new
date the height of tide above site datumwas expected to be only 24 ft. instead
of 26 ft. Further tests were therefore made with the standing ways lifted
vertically a distance corresponding to 2 ft. The length of ways below water was now considerably less than before, and the ship would have a longer run
and, therefore, a higher speed of entry. It was observed that the model tipped
outwards after leaving the way ends considerably more than with the 26 ft. tidesome 18 degrees in place of 12 degrees. It was, therefore, recommended
that the standing ways should be extended so as to give the same depth of water over the way ends as with the 26 ft. tide. Mr. Hughes went into this question on the site, and found the maximum possible extension was 15 feet,
and the corresponding length was added to the model ways and the programme
of tests repeated. Films of the wave on the beach were not taken on this
occasion, as the previous tests had shown any danger thereto be non-existent, and with 2 ft. lower tide the wave would not be expectedto reach so far up. Launching Speeds
The records of launching speeds against travelare plotted in Fig. 6, and the
maximum speeds, speed at entry and other data for both the 26 ft. arid 24 ft.
tides are given in Table 2. Curves of speed and acceleration to a base of time are shown in Fig. 7.
Angular Motion
As the pontoon left the ways ends, she tilted outwards from her static position on the ways, and as she became waterborne returned to the upright. The change in inclination with time was measured from the films, and curves of maximum values of the angle from the vertical for the experiments with 24 ft.
tide are shown in Fig. 8. The angle did not appear to depend on the speed of entry, since the curves of (i were not arranged in order of speed, but were apparently fortuitous, and only the maximum values are shown in Fig. 8. The maximum augular acceleration derived from the curve of maximum O
is also shown in Fig. 8. The concrete pontoon carried a very high steel structure
on top, reaching some fifty feet above the deck, and the acceleration forces
on this and consequently upon its anchorage to the deck were a source of some
anxiety; the values of maximum angular acceleration were of great value to the designers in this matter.
In the tests, the spaces between the wayswere made up to some five inches
below the top surface of the standing ways, that being the condition on the
site when the model tests were done. This was not a fixed condition, however, sand and shingle being continually removed or deposited throughout the year.
Some additional experiments were therefore made with the wood between the ways removed completely. The standing ways were then in water of some 40 ft. depth with only bracing struts between them. There was a much greater space for water to escape under the hull. The principal result was
TABLE 2Launching Speeds and other data for 26 ft. and 24 ft. Tides
Pontoon
has
travelled 220 jt.
425
Travel from rest until
maximum s peed is reachcd 1040
950
920
880
860
850
1120 1110 1100 1090 1070 1050 Minimum freeboard fronideck to top of wave
u during launch z About 4ft. u Just less It' thaP 4 ft. About 35 ft. 11 ,, ,. z -C.' ,, ,, o z 3.5 ft 3O ft.
Speed in ftísec. when
Tide height in feet
above First Speed Centre line of bilge is a site touches ,axi,u,n datum water
pontoon passes way ends
26 ft. 1455 1678 15 92 1780 9.4 1744 1830 to
98
1832 1914 1968 2013 22 00 22 80 24 ft. 1833 20 13 19 32 2070 100 2040 21 25 to 21 70 2260 107 2305 2430 2480 26 00 ,, About 40 ft. .' ul'i s, ,, Z (1i
460352 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
8 .Anah'sis of Launching Diagrams
Launching Speeth
For similar conditions of accelerating and retarding weights, the speed of entry was some 3 to 4 ft. ,sec. greater with the 24 ft. than with the 26 ft. tide,
due to the longer run from rest, as shown in Table 1. Once the model entered the water, she experienced very little further acceleration, because of the
resist-ance offered by the water, both in the form of head resistresist-ance on the leading
side and eddy resistance behind. Thus, with a speed of entry of 24 '8 ft. ¡sec.
on the 24 ft. tide, the maximum speed was only 26 ft. /sec.
The water resistance was greater the higher the speed of entry, and the deceleration was consequently greater also, with the result that with a 24 ft. tide the speed as the centre-line passed the way ends only varied from 10 to 10'7 feet/second, despite a difference of entry speed of from 18 '33 ft. ¡sec.
to 248 ft./sec.
This effect continued after she was afloat, and the finalspeed was the same whatever the initial speed of entry, being 4 6 ft. ¡sec. after
220 ft. travel on a 24 ft. tide. Thus the experiments indicated that although
the actual speed of entry on the pontoon might vary considerably with
tempera-ture and wind conditions of the day of the launch, the speeds passing the way ends and when afloat would be much the same in every case.
Angle of Tilt near Way Ends
As might be expected from the fact that the speed at the way ends was very
nearly the same for all speeds of entry, there was little variation in angle of tilt,
the maximum angle for the 24 ft. tide always being between 10 degrees and
12 degrees. The maximum angular acceleration during the tilt was O'65 degrees/sec. /sec.
TABLE 3Particulars of Wave on Far Side of River
Tide height in feet above site datum Speed of entry into water ft. Isecond
Time from release of pontoon to first wave reaching wall
on far bank in seconds
Maximum height of waves running up beach measured from still water
level in feet
Remarks
26 1455
416
2'5 Splashed over15 92 39.4
20
wall and some1744 37 . I 3.5 water spilled
18 32 362
40
over into gardens19 68 34.4
40
22' 00326
40
24 18'33407
About 40 ft. As above 19 389 2040 380 at top of feet 21 '70 362 23 '05 35 .3 2480 34.9SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING 353
If the problem of this launch is treated statically, there should be no such tilt, since there was enough water over the way ends to float the model off before reaching themthis was demonstrated by lowering the model SloWly down the ways. In the dynamic conditions of the launch it would appear that the tilting must be about the leading ends of the sliding ways, caused by the very high water resistance acting low down on the hull and the relatively high position of the centre of gravity.
(c) Wave Formation
When the model entered the water, a large "breaker" was formed along the whole length of her side. The minimum freeboard from the top of this breaker to the deck is shown in Table 1, its least value with a 24 ft. tide being
about three feet. With no "ground "between the standing ways, the breaker
was larger, the model tilted more, and a bulwark had to be erected along the deck edge. The top of the wave in this condition reached as far as 3 3 feet
above the deck edge. There was very little backwash up the berth behind the model as she left the waysagain a reassuring fact for the contractors, anxious for their plant on shore. As the model slowed up in the water, the breaker travelled across to the opposite bank, and the times and heights are shown in Table 3.
The maximum height above still-water level was some four feet, and some water spilled over the low walls into the garden.
9. The Ship Launch
In order to correlate the model and full-scale data, every endeavour was made to obtain complete records of the real pontoon launch. The travel and speed down the ways were measured by means of a wire attached to the pontoon and carried over a wheel on shore which made electrical contacts as it turned. In order to measure the angles of tilt, the freeboard at the greatest angle, and the wave on the far shore, records were taken on four
ciné cameras.
The launch duly took place on May 4th, 1943, in ideal weather, as near
"Tank" conditions as could be desired. The pontoon went off exactly to
programme, and the model predictions were almost completely fulfilled, as can be seen from the film which will be shown after the reading of the paper.
10. Comparison between Model Predictions and Actual Performance
Before comparing actual figures, it is necessary to mention one or two differences between the assumed and actual conditions.
The height of tide at the time of the launch was 23'-lO", i.e. very nearly the same as the 24'-O" used in the model experiments.
The GM was slightly greater than originally calculated, because some of the steel superstructure was incomplete. This would be expected
to reduce the angle of tilt on the ship.
The conditions on the far shore were rather different from those in the
model tests when the film was made. The initial height of water was 2'-2" lower, because in the model tests, when the ways were raised
to give the lower tide condition the model river bed and bank were not
altered because of constructional difficulties.
The surface of the
model and real beaches were of a different nature, as already stated.Also, the cross-section of the river in the model tests was made constant and equal to that along the centre-line of the launch. i.e., to the narrowest
part of the river, whereas in fact the river widens quite rapidly
im-mediately to each side of this line. In addition, the wave in the model
tests was confined by the tank walls within a water space equivalent to 400 ft. up and down river, or twice the length of the pontoon. Ali ax
354 SOME MODEL AND FULL SCALE EXPF.PIMENTS ON SIDE LAUNCHING
these differences would tend to decrease the height and penetration
of the wave on the shore. On the other hand, the height of the standing
ways surface above the intervening ground on the site varied from 18 in. to 30 in., as compared with the 5 in. in the model tests, and this would probably result in a larger initial breaker. On balance, it is
probably fair to say we should expect a slightly lower wave than
observed in the model tests.
The curve showing the speed and distance measurements made on the ship has been added to the predicted curves shown in Fig. ô. The general agreement in shape is remarkably good, and the final speed at the end of the
launch was almost exactly as predicted. It was actually reached a little earlier than expected, probably because the air resistance of the large superstructure-a most resistful superstructure-arrsuperstructure-angement of girders superstructure-and plsuperstructure-ates not on modelmsuperstructure-ade the deceleration more rapid. It will be seen that the actual ship curve lies a little below the curve for the model when launched freely on roller bearings, indicating that the coefficient of friction on the ship was somewhat higher than that on the model. The acceleration at the beginning of the ship launch actually corresponds to a coefficient of friction of 0 03, as compared with the figure of 0 025 given by the initial experiments with Russian tallow.
In Fig. 8 is shown the curve of heel for the ship,as measured from the
ciné record. The maximum value of 9 degrees was less than the maximum
for the model, 12 degrees, which would be expected from the larger GM on the shipthere was no opportunity of inclining the latter after the launch, as it was imperative that she be towed away at once.
The wave behaved much as expected, and just reached the garden wall. The gateways had been filled with sandbags and no water penetrated the
gardens. The following comparison brings out very clearly the close agreement
between the predicted and observed behaviour (Table 4).
TABLE 4Comparison Between Model Prediction and Actual
Behaviour of Pontoon
!tcPn
Height of tide Speed at entry
Time from release to bilge touching water
'Time from release to wave reaching far shore
Minimum freeboard from deck
to wave surface Distance reached by wave
Predicted from model experiments 24' 0 19-7 ft. ¡sec. 975 secs. 38-4 secs. 40 feet Splashed over wall
Observed on actual lazoich 23' 10" 197 ft. ¡sec. 10-8 secs. 37 secs. 20 feet Reached wall and
up it 6" to 8"
11. The Pier Head Pontoon launch at Marchwood
The second launch was of a pier head pontoon for the Mulberry Harbour, designed by L. G. Mouchel & Partners, Ltd., and constructed by Messrs.
SOME MODEL AND PULL SCALE EXPERIMENTS ON SIDE LAUNCHING 355
The outline shape is shown in Fig. 9. For the ship launch there were four standing ways, each twenty-four inches wide, all straight with a slope
of 1 in 16. The pressure on the grease was 1 75 tons per square foot.
Model Experiments
A model of the pontoon, N.P.L. No. 2381, was made in wood to a scale of 1/10th full size, being ballasted to the correct weight. As in the earlier launching experiments with model 2207, the model was launched on two standing ways only, these being in this case straight, with a slope of 1 in 16,
faced with steel strips. The model was fitted with two corresponding sliding ways carrying roller bearings, and with two dummy sliding ways so as to present
the correct scale head area for the ship in order that the comparable water
resistance would he experienced in each case. The space between the standing
ways was boarded in to give the correct level between ground and grease surface on the actual site, a matter of six inches. The details of the set-up
are shown in Fig. 9. After the first experiments had been made, it was found
that substantial cross-bracing was being fitted between the sliding ways, and further tests were done with this incorporated in the model ways (see Fig. 9). By varying the level of water in the Tank, different heights of tide above the site datum could be obtained.
The technique of drag weights and method of observation were generally similar to those described earlier in the paper. The pier head pontoon was to be launched into Southampton Water. There was thus no question of damage to the opposite shore, and observations on the wave created were not made, but the general disturbance- can be judged from Fig. 10.
The Ship Launches
The launches of two pier head pontoons at Marchwood were observed. The curves of distance run and the time were obtained in each case by the
method previously described.
Comparison between Model Predictions and Actual Performance
In Table 5 are given the principal results from the model launches with
three different heights of tideil' 0", 12' 3" and 14' 4". The experiments
with the last mentioned tide were made with both open and braced sliding
ways.
-The higher the tide, the less the travel before the sliding ways touch water, and hence the slower their speed of entrythe reduction being from 12 9 ft/sec. to 96 ft/sec. for the tide range 11' 0" to 14' 4". The maximum speed
was always recorded when the first bilge entered the waterfrom that point
onwards the model began to slow down under the effect of the water resistance
of the hull which was now added to that of the sliding ways. An important feature was the material reduction in speed of leaving the ways brought about by the higher tides, it being reduced from l2 15 ft/sec. with 11' 0" tide to
only 455 ft sec. with a 14' 4" tide, while after 320 feet of travel, the model being then well afloat, the respective speeds were 4 6 and 2 7 ft/sec. This suggested that if any of these units were to he launched on days which would
Wates, Ltd., at Marchwood, near Southampton. The pontoon was a concrete
structure of rectangular form, having the following principal dimensions
Length ... ... 80' 0"
Beam ... ... ... ... 56' 6"
Depth ... ... 15' 0"
Draught ... ... 6' 9"
356 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
give a 14' 4" tide, and the coefficient of friction was up due to cold weather or
other causes such as one of the ways not being true, there would be a danger of the pontoon sticking on the ways due to insufficient speed of entry. The recommendation was, therefore, made that in such a case the launch should be made some time before high water to ensure sufficient speed.
TABLE 5-Effect of Tide Height and Bracing of Wai's on Launching
of P.H.P. Model
(All figures given for ship
A comparison of the last two columns of Table 5 shows the effect of adding transverse bracing between the sliding ways. This bracing was only fitted to
the front end of the model ways to present approximately the correct head
area, but this would not exactly reproduce the ship condition where the bracing
was fitted throughout the length of the cradle. The effect on the ship launch would, therefore, be still greater than that given by the model experiments.
Naturally, up to the point where the sliding ways enter the water, the bracing has no effect. At all subsequent points it reduces the speedthe maximum
from 17 -8 to 16 3 feet/secondand the speed when clear of theways after 320 feet travel, from 2 7 to 2 2 feet/second.
The launches of two units at Marchwood were anended and ship data
ob-tained which are shown in Figs. li and 12 and Table 6. The tide heights
above site datum were 10 75 feet and li 25 feet for pontoons A and B
respec-tively. In Table 6 are shown the full-scale results and those predicted from
the model for an 11 0 ft. tide, both with and without the bracingbetween the
sliding ways.
lt will he noticed that in pontoon A there was a sudden slowing up at about 255 feet travel, with a violent deceleration. This was due to the fact that one
of the end standing ways was higher than the rest with the result that the pontoon held up at that side began to slew round. Any comparison with the
Tide height above
site datum Open Ways Braced Way
11' 0" 12' 3" 14' 4" 14' 4" When sliding ways
touch water
:-travel 83 ft. 63 ft. 30 ft. 30 ft.
speed 129 f/s. Il 'R f/s.
96 f/s.
96 f/s.
When first bilge enters water
travel 185 ft. 165 ft. 132 fr. 132 ft.
speed 193 f/s. 185 f/s. 178 f/s. 16-3 f/s.
When first bilge is over end of standing ways
:-travel 251 ft. 251 ft. 251 ft. 251 ft.
speed 1215 f/s. 101 f/s. 455 f/s. 443 f/s.
When ship is clear of standin wave
travel 320 ft. 320 ft. 320 ft. 320 ft.
SOME MODEL ANt) FULL SCALE EXPERIMENTS ON SIDE LAUNCHING 357
TABLE 6Comparison of Model and Ship Launches for P.H.P. Units
2Y
These are from model launch with no retarding weights in use, and unbraced ways. f Estimated from model launch with allowance for effect of partial bracing.
model is, therefore, useless beyond that point. It will be observed that
although both pontoons started off with much the same acceleration, the latter fell off much more quickly with B than with A once the cradle entered
the water. The only difference between the conditions on the two launches, as far as is known, was that the launch of B had been postponed for some
eleven days after the ways were greased, which may have increased the friction over the latter part of the run where the grease had been exposed to weathering
and the effects of sand. In Fig. il are shown the measured speeds for the
two pontoons and for the model when launched freely and with a small retarding weight. To this figure has been added a further curve showing the estimated
speed with braced ways. This has been obtained from the curve for the freely
launched model on the assumption that the percentage reduction in speed would be the same as observed for the model with and without bracing on a 14 feet 4 in. tide, and as set out in Table 5. Up to the point where the cradle entered the water, the ship launches agree very well with the estimate made
from the freely launched model. From that point onwards the ship launches
Item P.H.P. Pi-LP.
"B"
Model* ModeltTide htight above
datum 10-75 ft. 1l25 ft. 1100 ft. 1100 ft.
Height of fore-end of top of cradle above
standing ways 632 ft. 710 ft. 632 ft. 632 ft. Travel when sliding
ways reach water 87 ft. 79 ft. 83 ft. 83 ft.
Speed at this point l36 ft./sec. 126 ft./sec. 129 ft./sec. 129 ft./sec. (13-4 ft./sec.
at 83 ft. travel) Average acceleration to
this point 108 108 1-04 104
ft./sec./sec. ft./sec./sec. ft. /sec./sec. ft./sec.!sec. Average coefficient of
friction to this point 0029 0029 0-030 0030
Maximum speed 173 ft./sec. 150 ft./sec. 193 ft./sec. 175 ft./sec.
Travel at maximum
speed 153 ft. 153 ft. 185 ft. 185 ft.
Travel when bilge enters water
189 ft. 184 ft. 185 ft. 185 ft. Speed at this point 151 ft./sec. 138 ft."sec. 193 ft./sec. 175 ft./sec. Travel when bilge was
over end of standing
ways 251 ft. 251 ft. 251 ft. 251 ft.
358 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
were both slower arid the maximum speeds reached less than expected. As
stated earlier in this paragraph, the model ways were not truly representative of the ship ones as regards bracing, which would account for some of this
difference.
15.Conclusion
The launches described in this paper were ail of the sideways kind in which the standing ways were built out under the water, and the model or ship allowed to float off.
Model experiments can be used for investigating the probable behaviour
of a ship during sideways launching.
In the launch at Conway the model and ship arrangements were in every way identical, and the agreement in the shape of the various curves is
very good. Launching the model freely on roller bearings gave a speed of
entry just a little faster than that for the ship.
0 (3) Owing to the effects of temperature on the friction qualities of the grease, and the possible influence of wind and other external factors, the actual speed
of entry for the ship cannot be predicted accurately, and it is necessary to
cover a range of such speeds with the model by the use of artificial accelerating
or retarding devices.
The results for the pontoons agree wjth the model prediction up to
the point where the cradle enters the water. Beyond that point the ship launches were considerably slower.
In this cage the cradle on the model was in accordance with the original plans for the ship, but cross-bracing was added in the latter case before the
launch. It is believed that this is the principal cause of the differences in performance. Subsequently some bracing was added to the model cradle and
further experiments made. It was only fitted to the fore-end of the cradle, whereas that on the ship extended throughout, so that although the projected
head area was correct, the resistance of the bracing behind was not represented.
The bracing had the effect of slowing the launch down quite appreciably, as has been shown in the text, and it is concluded that in all such model tests in future the cradle must be made an exact replica of that for the ship.
The height of ground between the standing ways has an important influence, and must be correctly represented in the model arrangement.
Some values of coefficients of friction in side launches have been published in America and are reproduced in Fig. 13. The values observed
on the ship launches described here have been added.
The wave formation caused by the ship entering the water followed that predicted from the model experiments very closely.
The effect of height of tide in such launches can be gauged from the experiments with the model at three tides.
These showed that the higher the tide, the slower the speed at which the model passed the way ends, other things being the same.
16. Acknowledgments
The work described above was carried out in the Ship Division of the National Physical Laboratory on behalf of Director of Transportation (Tn.5), War Office, by whose permission this paper is published.
The Authors wish to record their thanks to Mr. Torys Hughes, Mr. Wood
of Mouchel and Partners, and to
the Contractors, Messrs. HollowayBrothers (London), and Wates, Ltd., for their great assistance in obtaining the full scale data.
The Authors also desire to acknowledge the assistance rendered by members
of the staff of the Ship and Physics Division of the N.P.L. in both the model
LH
2Z
SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING 359
Fig. 2
Fig. laSide Launch with Fixed Fig. lbSide Launch with Tilting
Ways and "Drop Off." Ways and "Drop Off."
Fz. icSide Launch.
Fig. idSide Launch. Fixed Ways360 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
I
Fig. 3Layout of Building Site at Conway. Model 2207.
a
ntp
CC
Some Model and Full Scale Experimentr
on Side Launching."
Paper by F. H. TODD, B.SC., PH.D., ASSOCIATE MEMBER, AND E. LAWS
LE VAllON _L
-e O' ! VW LX.'klNG LP ' Fig. 4Model 2207.TRAVEI FEET FOQ tt CURTE Fig. 7Model 2207 E . Q1 R ti
MODEL 2207 CONCRETE PONTOON
CLRVES O4 LAUNCHING UREEDU
ANO AECLRATION 24FEETTIDE HING SPEED
ut
o z 4 6SECURES Rif VELOCITY
o
r:
ii:
Plate XV
"Some Model and Full Scale Experiments on Side Launching."
25 TRAVEL N FEEr Fig. 6Model 2207
-FIG6 MODEL 2207 (cONcRETE PONTOOD)
SPEWS CLVES OF LALJNCHIFA3
-W.
-24 FEET TIDE - 1111__1 -oc or (tTRv cc
:T:
e 7 PE SE MODEL ,_____,ITi
IfiJJICH 51 POPI 00H MOO LA 20 n X o '5 V uu Io 5i
-362 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
Fig. 8Model 2207 (Concrete Pontoon)
Curves of Heel, Angular Velocity and Angular Acceleration 24 feet Tide. Ship Speed at Entry into Water: 1930 ft. per sec.
la
i.
jí
allai
A
WA
e 'SS 'S IL LAUN - TQQÑIII
I
ALIIIk
*.
1;
ii
4
3B
SOME MODEL AND FULL SCALE EXPERIMENTS ON SiDE LAUNCHING 363
Fig. 9
Fig. 11Pier Head Pontoons.
Curves of Launching Speeth.
o (l o o hi M o M M CM Mhi hi hi hi hi z o z CM hi hi M (D z
u.
..
u
u
u
rr uuuuÌu
__tuuuiui
__Hi!PiÍ'í
i.uL
Uil
¡it
uit-½U
HIUR
iiuuuit
rEAVft N FElT
TNt t»z
Fig. 12Pier Head Pontoons.
Curves of Launching Speeds and Acceleration
a' UI
I....
11h.. I III
I
i___
.1
III
uirn.1uuiuu
::iiuiu
UiUIIII
u Ivi
I.I 114
Ili
P
_____'IIE
t
ii
_____iiurniuun
111111
nu
Will
....lIlh'hI
UlII
.___
.
III
.
__
IM hlill
Ílii
I:
IH
Ihl.
__Ill.
__Ill.
. .III Uihl
IiirA
SIII1IIVA.i
SliliJIi
i_11111.
PllUlilllIll
.uI.0
iilUi
IUIDPd1I1II!1
Î
I
I
I
I
iii
.1
i
/
i..ì
III
il
i
lu
366 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING COEFFICIENT OF rNCrIa. Fig. 13. T
N
N
DISCUSSION ON "SOME MODEL AND FULL
SCALE EXPERIMENTS ON SIDE LAUNCHING"
The PRESIDENT (Mr. H. B. Robin Rowell, A.F.C., DL.):
We have listened to a very interesting paper. The Authors have made reference
to the different systems of launching ships sideways. lt is, of course, the fourth system (D) which is the system of launch-ing referred to in this paper, and I notice that the speeds attained are approximately
twice the speed normally attained by a vessel in the usual accepted manner of
end-on launching.
I do not appreciate the reason for the high speed or, in turn, the extent of the declivity. It is perhaps due to the geographical profile of the shipyard.
High speed, of course, is required for launching in the case of the other methods
where the tipping of the vessel at the
quay wall is of consequence.
I am interested that roller bearings had
to be resorted to for this experimental work. Tallow behaves strangely in small-scale experiments and it is interesting to note that some shipyards even today carry out experiments with weighted
launch logs to verify the quality of the
tallow supplied. I do not know upon what experimental characteristics they base their decisions.
Mr. W. R. G. WHITING, M.B.E., Menther
This account of unusual experiments I find fascinating on precisely the lines the President has suggested, and it is highly satisfactory to realize that success attended
these exploratory efforts, just as it did
those of the President.
There are two small points which occur to me: I am not quite clear why the angle of tilt should be greater if you have more water under the pontoon at the moment
of entry; in other words a higher tide.
Could Mr. Laws explain the factors under-lying that actual experience ? The other point is this : are the dimensional relations
which were employed to correlate the
model and the full-size launch similar
throughout ? In other words, do the
32
relationships between the model and the real hull follow the normal law of com-parison, or are there any special corrections which have to be made in sideways launching for the isolated wave front, the
pitching motion and the suction effect
of the shelving beach ?
Mr. W. MUCKLE, Member of Council:
I am interested to learn that attempts
which were made to use grease between.
the ways on the model proved failures.
At first sight it would appear that this is due to the fact that if everything is reduced to scale the pressures would be much less
in the model than in the full size, and
consequently this would affect the frictional force which is known to depend on pres-sure. It is noted, howcver, that the way area in the model was reduced to allow for this effect. It would appear then that the cause of the failure must be in the
grease itself, by not behaving in the same way when applied to small areas as it does when applied to large areas.
I note with satisfaction that the Authors state that the model experimental results
are remarkably well reproduced on the
full scale. This, however, is directly
opposed to the view expressed by Mr.
Nevita and Dr. Anderson at this Institution last year.f I would suggest that an ex-planation of this is that in the cases referred
to by Nevitt and Anderson there was
some motion of the fluid present which
depended on the static pressure which, of course, is not reproduced to scale in
the snodai. irs the present instance, however, it is merely wave motion which
is being dealt with and so long as the
speeds are adjusted to satisfy the Froude law of comparison then the wave system obtained with the model should be faith-fully reproduced in the full size.
Mr. R. J. W. RIJDKIN, Student:
Am I right in saying that Messrs. Jo1.a
Brown & Co. Ltd. carried out model
experiments on the Queen Mary before she was launched ? I am not certain whether they used grease or roller bearings. * Paper by F. H. Todd, B.Sc., Ph.D., Assoc. t" Two Aspects of the Dynamic Launching Member and E. Laws. See p. 345 ante, problem," N.E.C.Inst., Voi. 62.
Mr. E. C. B. CORLETT, Student: What was the actual coefficient of friction used in the experiment ? Was its
magnitude as compared to that of the ship liable to any scale effect?
VOTE OF THANKS
The PRESIDENT (Mr. H. B. Robin
Rowell, A.F.C., D.L.):
We are very much indebted to Dr.
Todd and particularly Mr. Laws forgiving us a most interesting paper and film.
It reminds me of many years ago, when at Armstrongs we had the privilege of buiId ing and engining some gun boats forthe Siamese. Some of our people as well as a Siamese potentate came to see the launch. As the ship curtsied into the
water, the old Siamese gentleman clapped his hands and shouted something which
on being interpreted was, "Have it up
and do it again." Unlike then, through the medium of films, this can now be done and we can study our mistakes again and again.
We should be careful of the results obtained from experiments suchas these. They can but be a broad pointerto show
us the way and we must take no notice
of the decimal points, for after all launching ships is a hazardous occupation and the materials used, that is, tituber and tallow,
are capable of fairly wide variation of
quality.
The vote svas carried with acclamation.
[CORRESPONDENCE
SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING D221
Mr. C. R. J. WOOD:
The launching of odd-shaped concrete craft is a matter which only occasionally requires attention, and it was a matter for satisfaction to those concerned to have at their disposal facilities for experiment at the National Physical Laboratory to which the Authors of this paper devoted such able attention. Their consequent forecasts of expected full-scale results
gave us much assurance and guidance,
arid were very fully justified, as this in-teresting paper demonstrates.
The reason for the use of the somewhat cumbersome wedge-shaped sliding ways for the P.H.P. may perhaps be explained. The pontoon was constructed horizontally, not following the declivity of the ways. Other vessels were similarly constructed
(this making for ease from the purely
constructional point of view) and
subse-Reply to the President
A high speed of entry is necessary to
ensure that the vessel will run sufficiently
far to clear the ways. With sideways launches of this type, where the ways are
carried right out under water, the
re-sistance experienced by the vessel is much higher than with the ordinary end
launch-ing, and there is a danger of the vessel
coming to rest ori the ways before being properly afloat unless she enters the water with sufficient speed. While we would
agree with the President that care is needed in the application of this work
to other launch problems, the comparison
between model and ship results is, we believe, sufficiently close to encourage the use of model experiments to in-vestigate launches which are in any way out of the ordinary. Various alternatives
can be quickly and easily tried in the
model, and from the results given in this paper there is every reason to believe that the comparative figures thus obtained
would give a true representation of the
behaviour of the ship under similar conditions.
Reply to Mr. Gorlett
The actual coefficient of friction at the beginning of the launch was 003 and O 029 in the case of the Conway and P.H.P. launches respectively. For the corresponding models, the values without any retarding weight were 0028 and 0030, showing that launching the model on roller bearings running on steel-plate ways gave nearly the same value for this
CORRESPONDENCE
quently tilted, by the use of oil jacks, to the desired declivity; but P.H.P. was too wide for this method to be adopted and it was, therefore, necessary to launch
the craft on wedge-shaped sliding ways. The exceptional height thus given to the leading edge of the sliding ways, coupled
with the tendency of the craft to start rotating (in plan) if the resistance to sliding should prove to be slightly greater
on one side than the other (a tendency
which is specially important in the case of broadside launching, though negligible
in end-on launching) made the cross bracing mentioned by the Author abso-lutely essential. Without it such a craft
would be in great danger of collapse. The retarding effect of the cross bracing is clearly demonstrated in the paper, and its effect upon the minimum length of run necessary is important.
AUTHORS' REPLY
" Sids Launching on the Great lakes." Soc. N.A. and Mar. E'zg., Vol. 50, 1942. coefficient as found in the full-scale launches.
Reply to Mr. Muckle
Like Mr. Muckle, we have been unable
to understand the failure to launch the
model on greased ways when the pressure
was made the same as on the ship. We
were led to use roller bearings by a process
of trial and error, and due to wartime
conditions it was not till some time later that the paper by J. H. Fahey* came into our possession. It was then very interest-ing to find that he had gone through the same process independently and come to the same final conclusion. He also found reasonable agreement between model and full-scale launches.
In regard to Mr. Muckle's reference to the paper before the Institution last year by Messrs. Nevitt and Anderson, the applicability of model launches to the full scale is only referred to very
briefly in the authors' reply to Professor Burrill. They apparently had no in-formation nor had they any idea as to the discrepancies, and in effect agreed to suspend judgment until they had more
first-hand knowledge of such work. Reply to Mr. Rudkin
Mr. Rudkin is quite correct in saying
that Messrs. John Brown & Co,. Ltd. carried out model experiments on the
D222 SOME MODEL AND FULL SCALE EXPERIMENTS ON SIDE LAUNCHING
Queen Mary. The launching
arrange-ments were described in a paper by Dr. ÏvlcNeiIl before the Institution of Naval Architects in 1935,* but he gave no details of how the model was launched, merely
saying that the velocity of the model
was regulated so as 'jto correspond with the predicted velocity in the actual launch." Incidentally, the ship predictions based on coefficients and measurements made on the model were very closely realized in practice, showing the value of model work for end-on as well as side launches. Reply to Mr. Whiting
No special dimensional relations were used in applying the model figures to the ship-model. Dimensions were scaled up directly, and model velocities and times increased as the square root of the linear
scale when predicting the ship figures. Launch of the Quadruple-Screw Turbine Steamer Queen Mary," I.N.A., Vol. 77.
s
We have no explanation at the moment of why the deep water between the ways increased the outward angle of tilt. Had
the experiments not been done under
severe pressure of work during a critical phase of the war, this feature would have
been pursued, but there was no time to
do so under the conditions then existing. It should be noted that the increased angle
of tilt was produced by removing the "ground" between the ways. This is
not the same as a higher tide. Reply ro Mr. Wood
It is most interesting to hear from Mr.
Wood the reasons which led to the
in-troduction of the cross bracing between the ways. The results of any slewing are certainly likely to be much more serious in broadside launches, and such mishaps are by no means unknown. We are glad
to have Mr. Wood's assurance that our
work was useful to him and fully justified the use of model experiments in solving a new problem.