f - - - -

rI

I RIJKSWATERSTAAT,i

: COMMUNICATIONS!

No. 21

PUSH TOWS IN CANALS

DY

IR. J. KOSTER

1975

(2B 2032,

21

RIJKSW A TERSTAA T COMMUNICA TrONS

PUSH TOWS IN CANALS

by

IR. J. KOSTER

Delft Hydrau1ics Laboratory(up to 1-1-1975)

Government Publishing Office - The Hague 1975

All correspondence should be addressed to

RIJKSWATERSTAAT

DIRECTIE WATERHUISHOUDING EN WATERBEWEGING

THE HAGUE - NETHERLANDS

The views in this article are the author's own.

ISBN 90 12 00864 8

Contents

page

5

8

8

811

11

12

12

12

14

15

16

16

1718

1919202228

32323234353638

39

2

2.1

2.2

2.32.4

33.13.23.33.4

4

4.1

4.24.3

55.15.25.35.4

66.16.26.36.46.56.6

Introduction

Prototype measurements

Purpose of the studyNavigation habitsTraffic measurementsMeasurements with a push tow

Study of scale effectsIntroductionDefinition of the problemMeasurement resultsConcIusions

Influence of the canal cross-section on the directional stability of thepush towIntroductionMeasurement resultsConcIusions

Overtaking manoeuvresIntroductionSimulation of prototype collisionsPhenomena during overtakingRecommended canal section

Special studiesEffects on the canal sectionSide-windScheldt-Rhine CanaI/Eendracht bendAmsterdam-Rhine Canal/Demka bend and Goyer bridgeBeuningen outlet structureMaarssen intake structure

Figures

3

1 Introduction

This publication summarizes the research carried out in the Netherlands in the 1960sinto the dimensions required for the cross-section of ship canals if push tows are tonavigate safely with other types of inland shipping.

The research was conducted by a working party set up by the Director-General of theDepartment of Water Control and Public Works (Rijkswaterstaat); in addition to theappropriate Rijkswaterstaat services, the Delft Hydraulics Laboratory, the Nether­lands Ship Model Basin in Wageningen and the Department of Naval Architecture ofDelft University of Technology were represented on this working party.

The necessary cross-section dimensions were determined by the space required for thedifferent traffk situations which must be possible without danger. The traffk situationrequiring the greatest amount of space - taken as the key traffk situation - involvesthree large vessels: one is overtaking another, while a third is sailing in the oppositedirection.

At first sight it would seem that trafik problems in fairways can be approached bymethods similar to those used in the case of roads, i.e. by allocating a lane of a givenwidth to each of the vessels involved. This method has been used in some previousstudies. Closer consideration shows, however, that it is unsatisfactory because in afairway not only the vessel but also the supporting environment is in motion. As aresult, the transverse movements of a vessel are more difficult to control than those ofa vehicle on a highway whereas in addition, the different vessels infiuence one another.

As long as a single ship is sailing in the axis of a straight canal with a prismaticcross-section, it is possible to derive approximate theoretica1 values for at least theprincipal characteristic parameters of the water movement (drop of the water leveland reverse flow). In this case sufficient data is also available from practical measure­ments and laboratory tests to determine with reasonable accuracy the deviations fromthe theoretical results as a consequence of the approximations used.

However, when the ship is not navigating in the canal axis and in particular whenseveral ships are present alongside each other in the canal simultaneously, the watermovement becomes too complicated for a theoretical approximation. It is thereforestill impossibIe to obtain rules for the design of a canal cross-section by theoreticalmeans. Satisfactory data for this purpose is also not available in literature.

5

U ntil rcccnLly this was not a serIO liS prohlem occnuse- widc k..nmvlcdgc had beenacquired through long experience of traditional inland navigaLion (towed barges anoself-propelled ships) in which conditions (lypes of ship. dimensions and speeds oftra vel) changed only gradually (photo I). With the introduction of push tows. thisknowlcdge gaincd from cxperiencc ccased to be sufficÎent. Ol only are the push towsconsidcrably larger than the other types or \'essel. bUL bccausc or their different shape.especially of Ihe bow. lhe waler movemenl crealed by them has markedly different

characleristics (photo 2).

It \\as thererorc ncccssary to gain a beller insight Înto this problem through modeltests. Direct simulation (\\ith rree-moving model ships manocU\Ting undcr realisticconditions) of the passing manoeuvres liable to be encountered in practice on thecanal. formed all important part of the study. Serore this sinllllation could be earriedout, it wac;, however. necessary to obtain more detailed physical information on thcphcnamcna. at least in qualitativc tcrms. in spccially designed series of tests.

This was useful ror a meaningflll interpretation of the passing tests and also necessaryto determine the possible influence of seale effects. For both these reasons it was alsodesirabic la do thc most accurate passible prototype measlirements. Tile contacts

Piloto I. Tradilional navigation changed vcry gradually ~o lhat until recentl)' canal designcould be based on long experience.

6

r'hoto 2. The push tow is not only considcrably largcr lhan 01 her vessels bul also different inshape. cspecially at Ihe bow.

established in this connexÎon wilh the professionLIl circlcs directly cOllcerned providcdvaluable guidance.

A few separate series ol" tests conccntrated on special circumSlanccs sueh as sidc-winds,eallal bends and constrictions. A funhcr general study was made of the enccts oncanal banks of the water movement callscd by shipping. Some specific studies werea1so made of the Aow distribution charactcrisLies of intake and out let structuressiluated along thc fairways for \Valer management purpose. Thc aim \vas to limit asrar as po~sible lht.: hindrancc \\ hieh might be c.llIsed w shipping. and in panieularpush tow navigrttion. by slleh local1y-gcncrated cross-curreIllS.

Thc sludy was conducled in lhe De Voorst LaboralOry of lhc Delft HydraulicsLaboraLOry. Thc rcsults have alre<Jdy been applicd 10 the design of the widencdAmsterdam-Rhine Canal. thc Schcldl-Rhine Canal and the Hartel Canal (sechgurc I).

Thc rcsulto;; or thc slUdy are set out in detail in reports comaÎned in four annexes tothe final repon or lhe working party. These annexes - in DU1ch arc obtainablefrom the correspondence addrcss.

7

2 Prototype measurements

2.1 Purpose of the study

The main purpose of the prototype measurements and tests was to obtain adequateknowledge of the complex relationship between the innumerable factors which areimportant in determining the conduct of barge-skippers in face of traffic situations.This information was necessary to define the structure of the model tests and tointerpret the model measurement results.

The boundary conditions for the model study were also determined on the basis ofprototype measurements; these conditions included dimensions, draughts, capacities,navigation speeds of the ships, traffic situations encountered and the correspondingmanoeuvres: evasive action, adjusting navigation speeds, distances from othervessels and from the banks.

The prototype measurements also provided data for calibration of the model, inparticular calibration of the behaviour of the ship and the helmsman under statie anddynamic interference between the ship and canal. Here, the course angle in relationto the axis of the canal and the rudder angle - in particular the mean value andstandard deviation as a function of time - are important parameters.

Finally, the prototype measurements together with model tests on different scales,provided important information on scale effects in nautical model tests.

2.2 Navigation habits

To gain familiarity with the navigation habits of skippers on inland canals, a numberof voyages were made on board various inland waterway vessels. The experiencegained in this way was of great importance for accurate manoeuvres with the model.

For interpretation of the model results, it was also necessary to know what skippersconsidered an acceptable pattern of passing manoeuvres under practical conditions(photos 3 and 4).

Photographs 5 and 6 clearly show that dangerous situations may arise. The fact thatthese passing manoeuvres did not result in a collision was due to the goodmanoeuvrability which motor vessels generally have, unlike tow barges.

8

rholo 3. PU'ih tow consisling of onc barge rneets a motor ~hip at a 'normar distance.

Photo 4. 1olor ')hip o"crtakes a singlc·bargc push 10\\ "ith normal clearance.

9

Photo 5. Single·bargc push tow overtaken by an unIaden fast ship. with two oncoming 0\ er­taking vessels.

Photo 6. The space on eilher side of thc raS1 \esse! cannol be a norm for canal design.

10

2.3 Traffic measurements

The traffic survey conducted in 1964 on the Amsterdam-Rhine Canal gave a clearpicture of the overall traffk pattern. An analysis of the inland waterway ffeet interms of load capacity and distribution of navigation speeds shows that ships of2,000 tons are sometimes encountered and that speeds range from less than 7 kph tomore than IS kph (see figures 2 and 3).

A more detailed analysis ofthe results enabled the navigation speeds to be determinedat which the probability of being involved in an overtaking manoeuvre (overtaking orbeing overtaken) was smallest. This speed, which is between 10 and 13 kph, could berecommended for push tows (see figure 4).

2.4 Measurements with a push tow

The characteristics of a push tow consisting of 4 'Europe-I' barges (E I-barges), witha draught of 2.86 m, were measured in a section of the Amsterdam-Rhine Canal witha depth of 4.20 mand a width at the water level and bottom of 72 and 50 m res­pective1y.

These prototype measurements concentrated on navigation at constant speed andconstant distance from the banks, as these conditions could be reproduced relative1yeasily in the model.

The measurement results showed that the drift angle and especially the rudder anglefluctuated around a state of equilibrium (see figure 5). In genera1 the state ofequilibrium was not equal to zero.

11

3 Study of scale effects

3.1 Introduction

lt was necessary to determine the optimum scale for the projected model study. Onthe one hand the model must be as large as possible in view of scale effects in thehydrodynamic phenomena. (The hydrodynamic phenomena of resistance, propulsionand rudder action may be affected by scale effects as a consequence of an excessivelylow Reynolds number. There mayalso be question of scale effects in the behaviourof a helmsman in the model, for example in perceiving the ship's movements in casethe time scale is too small). On the other hand, building and operating costs make itcheaper to use a small model. Here too there is, however, a lower limit becausesmaller models place higher requirements on technical equipment if a given level ofaccuracy is to be maintained.

The scale effect study concentrated on the equilibrium rudder angle and drift angleof a push tow sailing along a canal bank and on the rudder action and resistancejpropulsion of a push tow. This study was carried out jointly by the Netherlands ShipModel Basin in Wageningen, the Naval Architecture Laboratory of Delft Universityof Technology and the Delft Hydraulics Laboratory.

No study was made of the scale effect in the behaviour of the model helmsman.Experience with previous model studies suggested that this scale effect will be slight,provided that the model helmsman is able to steer from the ship itself and has a lineof sight corresponding to that of the helmsman of the prototype. Investigation of thisproblem would not have fallen within the terms of reference of the working party.

The study was conducted with a push tow of 2 x 2 E I-barges and a pusher tug witha power of 2 x 750 hp (see figure 6). Models on three scales were used, i.e. 1: 40,I: 25 and I: 10.5. Together with the prototype tests, a wide range of possible scaleswas therefore covered.

3.2 Definition of the problem

In a model the physical parameters are reproduced in such a way that certaincharacteristics are the same for the prototype and model; in this study they were :

12

2the Froude number = ~ for accurate reproduction of inertia effects;

gl

vi11 the Reynolds number = - for accurate reproduction of the influence of

. . VVISCOSlty.

In these expressions vIgv

speedlengthacceleration of gravitykinematic viscosity

Since g and vare identical for the model and prototype, the following scale laws applyrespectively:

There is of course a wide discrepancy between laws i and ii. If n v is introduced on thebasis of an identical Froude number it follows from ii that the Reynolds number willbe a factor n3~2 too low. This leads to differences between the model and real situation.These differences are defined as the scale effect.

A push tow sailing parallel to a canal bank without drift angle and with its rudder setcentrally will be exposed to lateral farces:- The water levels on either side of the push tow will not be the same. This leads to

a difference in hydrostatic pressure.- The water under the keel will not flow solely in the longitudinal direction. The

frictional resistance therefore has a component directed athwartships (see figure 7).

The resultant of these lateral farces -- referred to below as the canal bank force - isperpendicular to the ship's axis or an extension of that axis.

Two reaction forces are needed to establish equilibrium with the canal bank force.One reaction force (drift force) occurs because there is a drift angle between the pushtow and its direction of travel. The other reaction force (rudder force) is caused bydeflection of the rudder.

The system is c1early defined sa that there is one value for the rudder angle and at thesame time one value for the drift angle at which the farces on the push tow are in astate of equilibrium. These values are known as the equilibrium rudder angle and theequilibrium drift angle.

13

It is known from the study that in a state of equilibrium the rudder is always directedtowards the bank, while the equilibrium drift angle, dependent on a number offactorssuch as the bow shape, water depth and distance from the bank, may be eitherpositive or negative.

The force diagrams in figure 8 show that this is due to displacement of the point atwhich the canal bank force acts. The equilibrium rudder angle and the equilibriumdrift angle are therefore largely dependent on the forces exerted on the push tow bythe surrounding water.

Since almast all the farces are caused by phenomena in which viscosity plays a part,both the rudder action and the effect of the drift angle on the interplay of farces maybe influenced by scale effects. As aresuit there may be differences between the sizeof the equilibrium rudder angle and the equilibrium drift angle in the model andprototype.Some phenomena will counteract each other, so that it is impossible to predictwhether the rudder angle and drift ang1e in the model will be larger or smaller thanin the prototype.

3.3 Measurement results

Some measurement results obtained iu the study of the equilibrium rudder angle anddrift angle of a push tow sailing along a cana1 bank are shown in figures 9 to 11.Figures 9 and 10 compare the mean rudder ang1e and drift ang1e during the differentprototype measurements with the corresponding values for a free-moving model on asca1e of 1: 25. Although the results coincide reasonab1y weil, it is significant that theprototype results show a wide scatter.Figure 11 shows the results for push tows held in a fixed position in the lateraldirection, on different sca1es. With reference to the holding farces, the figure showson a graph the rudder and drift ang1es at which the 1aten~.l farces are in equilibrium.As the measurement accuracy is 1 to 1.5 tons (prototype), the measured differences arenot significant.

Possible scale effects in the state of equilibrium when sailing along a canal bank areso small in relation to the measurement accuracy and scatter of the measurementresults that no sca1e effects cou1d be shown in the equilibrium rudder and drift angles.

The study of the rudder action involved measurement of the transverse farces as afunction of the rudder angle of push tows held in a fixed position in the lateraldirection on different scales (see figure 12). Taking into account the measurementaccuracy referred to earlier, there does not appear to be any scale effect here either.The study of scale effects on resistancefpropu1sion - expressed in terms of propellerspeed - also showed that the scale effect was so small that, for a variety of reasons,it could not be quantified.

14

3.4 Conclusions

On the basis of these measurement results and the considerations set out in theintroduction, it was decided to continue the study on a scale of I :25. The argumentthat the push tow and the tow barges used in this case would then be large enoughin the model to allow room for a helmsman was the determining factor.

15

4 Inftuence of the canal cross-section on the directional stabilit)'of the push tow

4.1 Introductioll

In the second phase of thc preparatory sludy a large numbcr of tests weTe carricd out

to determinc thc rclatiol1ship betwecn dimcnsions of the canal seclion and theirinf1uencc on the bchaviouT of a push tQW. A systemalÎc sludy ofthis kind had not beencarricd out bcfore. although the results arc vcry important for (he adequate design

of a cunni lIsed by push 10\\5. Thc depth. width and slope configuration of thc canal,and the propeller speed of tlle pu~her {lig. \\ere \ariable boundary conditions ofthe model. Thc range of variatioll of these parameters is such that the results areapplicablc lo \\idely difTering cases.

A push to\\ \\ith automatic steering was uscd in Ihis study (pholo 7). The slecring

characlerislics of this automatic systcm cao bc vuried in such a \\ay that the perform­ance of bath an unskillcd and idcal helmsman can for examplc bc simulatcd. Thc use

Photo 7. Alilomalically·stccrcd push low during stud)' of dircctional stability in different canals,scalc I : 25.

16

of an automatic system of this kind is important when the steering characteristicsmust be identical in all the tests of a series.InitiaJly the comparison parameters used in this study were derived from the averagerudder angle and the average drift angJe of a push tow encountered when the push towsails close to the axis of the canal. The relevant literature suggests that these para­meters are the most commonly used in studies of this kind. These parameters do,however, have the drawback of disregarding the non-steady phenomena whichare particularly important during navigation. It is precisely the extent to which theship and its rudder fluctuate periodicaJly around their mean position which gives anindication of steerability.In the remainder ofthe study, the parameters were therefore adapted more accuratelyto the dynamic behaviour of the ship.

4.2 Measurement results

Tbc mean rudder and drift angJes were determined from observations on a free­moving push tow on a scale of I :25. The number of voyages (at least 4) and the lengthof each voyage (approx. 4 km for the prototype) were large enough to determine themean value accurately. These averages were determined for a number of differentdistances between the push tow and the canal axis.

Figures 13 and 14 indicate as examples the results for rectangular canal sections witha width of 100 mand water depths varying from 3.80 to 7.40 m. These results showthat direct comparison of the cross-sections was not possible. The absolute values ofthe rudder angle and drift angle were small in all cases and always permissible. Inaddition, the rudder angle characteristic appeared to be much the same in all canalsections, regardless of the variations in the section shape. Only the drift angle charac­teristic showed noticeable variations, depending on the canal section (see figure 14).

The reasoning foJlowed in the interpretation is therefore based on the characteristicof the drift angle at an increasing distance [rom the canal axis. The drift angle is aconsequence of the canal bank suction acting on the push tow. A steep gradient inthe drift angle characteristic therefore signifies strong local variations in the interplayof forces, which may be experienced as an obstacle by the helmsmen. In this way itwas possible to formulate a number of reql'irements with which a canal section mustcomply.

As the study progressed, emphasis was also placed on the extent to which the rudderand drift angles fluctuate around their state of equilibrium during navigation. Infigure 15, the behaviour of the automatic pilot in the model is compared with thebehaviour of a skilled helmsman during one of the prototype measurements. lt is to

17

he expected that the differences ohserved between the model and prototype will haveonly a slight influence on the results of the study.The standard deviations ofthe rudder angle and drift angle appear to be comparativelylow under practically all conditions They did therefore not appear to be a sufficientlysensitive parameter for the influence of the canal section on the directional stabilityof the push tow. As an example, figure 16 gives an impression of the measurementresults.

4.3 Conclusions

Together with the previous study, the tests led to the following conclusions :

The canal water depth must not he less than about 5 m, corresponding to 1.5 timesthe draught of the largest vessel.The shape of the underwater slopes has no noticeable influence on the directionalstability of the push tow.If the water depth immediately in front of the canal bank is not less than 1.1 timesthe draught of the largest vessel, the navigable width and the directional stabilityare no smaller than in a completely rectangular canal section.

18

5 Overtaking manoeuvres

5.1 Introduction

The principal aim of the model study of push tow navigation in canals is to determinethe cross-section of canals intended partly for push tow navigation. The cross-sectionswere studied under two conditions :

- A push tow navigates singly on the canal (see chapter 4).- A push tow and other ships pass each other.

in the first study the cross-sections were compared with the aid of a parameterderived from the rudder angle or drift angle encountered when the push tow sails insuccession through canals with different cross-sections.

A similar methad of study was used in designing the North Sea Canal (Delft Hy­draulics Laboratory report M 726), the North Sea-Baltic Canal (Hansa No. 36/38,2-9-'52), the Panama Canal (Lea, C.A. and Bowers, C.E.; Panama Canal, ShipPerformance in Restricted Channels; Proc. ASCE 1948) and the Suez Canal.

In busy canals it must be possible for two ships of the largest permissible dimensionsto sail past each other in opposite directions ; if the traffk density makes it desirabieeven more than two traffic lanes may be needed for sufficient space for oncomingvessels during an overtaking manoeuvre. The traffic situation which is considereddecisive for a particular canal wiJl depend on the desired capacity of that canal.

The study was based on a traffk density which makes it desirabie to have suffi­cient space for oncoming vessels during an overtaking manoeuvre (three navigationlanes). In addition, it must be possible for the ships to travel sufficiently fast to makeeconomic use of their engine power. This means for example that a push tow shouldnot have to reduce its speed to less than approx. 10 kph.

The distribution of vessel types makes it unnecessary to allow for all conceivablecombinations. When the daily traffic in bath directions is estimated at 5 to 10 pushtows for example and 150 to 200 other vessels, allowing also for the high level ofuniformity in the dimensions and speed of push tows, overtaking of two push towscan be ruled out. The probability that the three traffic lanes wiJl be used simultaneouslyby two push tows and a normal tow can also be considered very low.

19

The preliminary study showed that the traffic situation which creates the greatestdifficulties is that in which a large ship with poor steering characteristics is overtakenby another large ship. For this reason, the decisive situation was chosen as that inwhich a normal tow is overtaken by a push tow with 2 x 2 wide barges, while a motorvessel is sailing in the opposite direction (see figure 6). Speeds of 7 kph for the normaltow, 10 kph for the push tow and 15 kph for the motor vessel were chosen. The aimof the study was to allow the manoeuvre to take place without difficulty under theseconditions, naturallyon the assumption that the skippers complied with the require­ments of good seamanship, in particular by reducing speed where necessary, andallowing each other enough room. At the speeds indicated, the overtaking manoeuvretakes 13 to 14 minutes and continues over a distance of about 2.1 km. If there is noroom for oncoming traffic, the overtaking manoeuvre can only be begun if the closestoncoming vessel is more than 5.5 km away. This shows that three lane navigation isnecessary when traffic is dense.

The manoeuvring tests were based on the cIosest possible approximation to the realcharacteristics of shipping. For this purpose, the tests were designed and carried outin close contact with canal ship operators (photos 10 to 15).

The model tests were carried out with free-moving ships. During an overtakingmanoeuvre the successive positions of the ships were recorded photographically(photo 16). A parameter was then derived to assess the safety of the decisive trafficsituation; this parameter is characteristic of the relevant canal section. lt is alsonecessary to determine what values the parameter must have for the canal to besufficiently safe.

The traffic situations in the model can also be assessed visually to determine theirsafety.An expert eye-witness will obtain a good impression during observation of the shipmovements, to the extent that safety diminishes as the ships are closer together.A film was therefore made of passing manoeuvres in the principal canal sections (seephotos 8 and 9). Safety in different canal sections can then be compared withoutrelying excessively on the memory of observers. At the same time persons who werenot able to follow the progress of the tests from day to day can gain an impressionthrough this film of the course of the manoeuvres and the differences in the canalsection dimensions.

5.2 Simulation of prototype collisions

A number of prototype collisions were simulated in the model in order to investigatewhether a situation which in practice appeared to lead to collisions because of a high

20

Photo 8.

l.lhutQ 9. Dccisi\C traflk ::.Îluation in thc model. Passing manOCUHes are filmed rrom the instru­ment troiIe) Ira\elling bchind the 10\\.

21

speed of travel or an insufficiently wide passing distance, also made an unsafe im­pression in the model. These tests were repeated several times in the model and thephenomena were found to be reproducable to a high degree; they also coincided withthe prototype observations.

An example of a collision later on simulated in the model is described in the mi­nutes of the Netherlands Water Police (see figure 17).

5.3 Phenomena during overtaking

As already indicated the behaviour of the tow barge is the most critical in the decisivetraflic situation. The push tow and oncoming vessels maintain their course weilwithout deviating significantly from it and remain practically straight in the canal.The tow barge cannot, however, hold its course when it is overtaken by the push tow.The forces which push the tow barge olf course are mainly due to the water movementprevailing around the push tow.

In front of the push tow a slight increase in water level occurs and the reverse flow,generated by the push tow, has curved flow lines there. Because of this phenomenona force acts on the tow barge sideways of the bow of the push tow; this force is exertedtowards the starboard bank and alfects the rear of the tow barge first. The Jatter thenslews over towards the bank while the forward part ofthe barge moves away from thebank. As the push tow continues its overtaking manoeuvre, the repelling force isdisplaced towards the bow of the tow barge. In this situation, the tow barge is angledin relation to the bank but still moves roughly parallel to it. MeanwhiJe the rear partof the tow barge drops into the Jower water area which begins just behind the bowof the push tow. As aresuit the barge train is slowed down considerably.

When the whole tow barge is alongside the push tow the latter has little influence onit. The tow barge then has a short time to adjust its position before the most criticalpart of the overtaking manoeuvre begins. The water-level drop is quickJy made upagain to the rear of the push tow barges, so that the tow barge is acceJerated. Thisphenomenon continues untiJ the push tow has overtaken completely. The towingmotor vessel must then try to keep the tow line taut by increasing its speed. lf this isnot possible, the tow barge wiJl close up on it and will be less easy to steer becausethe tow line will no longer draw the bow in the correct direction.

At the same time the tow barge will be exposed to astrong suction force towards thepush tow. This lateraJ force is also a consequence of the reverse flow directed towardsthe psuh tow at the rear of the push tow barges and to alesser extent at the rear ofthe pusher tug. This suction force also acts first on the rear of the tow barge whichtherefore slews over towards the bank. If this angle is large enough the tow barge willnot be drawn far olf course but will move parallel to the bank.

22

Photo 10.

Photo 11. Steering the model push IOW compared \\lIh protot~pe. There is no room ror themodel helmsman in the pusher tug.

23

When the rear of the push tow has just passed the tow barge after overtaking, it isstill possible that the bow ofthe tow barge will be drawn over so strongly that the towbarge veers off course and slews over to port in the cana!. A collision with the pushtow is no longer possible at this stage but the tow barge may collide with an oncomingvessel. In practice, it is also possibie in this situation for the tow barge to collide withanother ship overtaking it after the push tow did. This course ofevents in an overtakingmanoeuvre, like the occurrence of a collision, must naturally be regarded as imper­missible.

The phenomena described above can be illustrated by an example. This concerns thebehaviour of the rearmost tow barge in a train of two barges and a motor vesselwhich is overtaken by a wide push tow. The canni is treated in this case as a two­lane fairway so that the push tow can move well over to port. This overtaking ma­noeuvre was measured 16 times in all.

The test in which the tow barge requires the greatest amount of space is illustratedcompletely in figure 18. This figure also defines the concepts of 'navigation lanewidth' and 'space adjacent to the navigation lane'. The navigation lane width is thewidth between two lines in between which are enclosed all the different successivepositions. The traffic lane width is the width of a canal lane.

All the measurements in this series showed much the same characteristics. It is there­fore reasonable to combine the observations into a mean value to enable the speedsand accelerations to be derived with greater accuracy by differentiation (see figure 19).

Although the phenomena followed the same pattern in all the tests as a function oftime, the values reached in the different tests may differ considerably. This is illustratedon the basis of the measurement results for a trapezoidal canal section with a waterdepth of 5 mand bottom and surface widths of 115 and 155 m respectively (see figures20 to 29). Figures 20 to 23 show that the model helmsmen of the tow barge and pushtow did not 'leam' during the series of 55 tests. Good and less good voyages occurredat both the beginning and end of the test series. Because of their experience of steeringmodel ships the maximum performance of the model helmsmen had already beenreached before the beginning of the measurements. The learning effect thus did notinterfere with the measurement results.

In the 3rd and 30th voyages the distance between the tow barge and push tow wasnegative (see figure 22). This means that after passing the rear of the push tow, thetow barge moved over partly behind the push tow. In voyage 3 this resulted in acollision with the push tow owing to the fact that on this voyage the tow barge hada very broad navigation lane width (see figure 20); the cause was not an excessivedistance from the canal bank (see figure 21) or between the push tow and the canalaxis (see figure 23).

24

Pholo 12 .

•PllOlo 13. Steering a model tO\~ barge comparcd \\Îth the prototype. ln:.u:ad of a tlig. a RI-IK

ship is uscd In the model steered by lhc pcrson at thc froll!. Thc person at thc back'iteer~ thc IO\~ b~trgc.

25

The model helmsman of the push tow is generally able to navigate in the centre ofthe canal with a good measure of accuracy. His instructions were to keep the starboardside of the push tow in line with the canal axis. He had no visible means of guidance.The ends of the model were hidden for this purpose by black plastic foil. The onlyguidance was provided by the line of the canal banks.

Whenever the push tow was too far over on the starboard side of the canal accordingto the personal impression of the tow barge skipper, the helmsman of the push towwas 'reprimanded'. This may explain why, after the 3rd and 30th voyages, a numberof voyages were navigated too far over on the port side (see figure 23).

The last three voyages of the push tow were made much (average about lOm) toofar over to the starboard side of the canal. This may have been caused by a lack ofconcentration on the part of the helmsman of the push tow as a result of fatigue orreduced motivation. In these 3 voyages the distance between the tow barge and thepush tow was no less than 12 m so that the tow barge skippers had little reason tocorrect the behaviour of the push tow helmsman.

In order to investigate whether the distance between the push tow and the canal axisinfluences the navigation lane width of the tow barge these values are compared infigure 24. There seems to be no influence.

lt is also important to know whether the navigation lane width of the tow barge isdependent on the distance between the latter and the bank (and push tow). For thispurpose figure 25 shows for each voyage the largest and smallest distance observedbetween the tow barge and the bank (measured at the plane of the keel) expressedagainst the navigation lane width. For smal1 navigation lane widths « 36 m) theminimum distance is 10 to 20 m. With a large width (> 36 m) the tow barge always(7 cases) comes within less than 10 m of the bank. In 4 cases the distance was evenless than 5 m. lt may therefore be concluded that the traffic lane required for the towbarge need not be any greater than the navigation lane width which is still conidseredpermissible.

The above observations show that the scatter in the measurement results is considera­bie; this cannot be ascribed to a dispersion ofthe material boundary conditions. The solecause must lie in the uncertain behaviour of the model helmsmen. Since the modelhelmsmen had the intention of steering the best possible course in each passing test,they gained the impression that the progress ofthe passing manoeuvre was determinedby chance. By analysing in greater detail the frequency distribution of the scatter inthe results, it is possible to determine the probability of the tow barge and push towmoving out of their traffic lanes.

26

Photo 14.

"how 15. \1odel ra,~ing manOCUHC. Thc push 10\\ and 10\\ barge \\ere sleered bye"perienced~kipper~ "ho (lClermined Ihe respCCII\C dlslanccs iOllliti\cl~.

27

The eumulative frequeney distribution of the traffie lane widths used by the push towand tow barge is shown in figure 26. The measurement results eontained in figures 20and 23 are used again here. The navigation lane width of the push tow was assumedto be 1.1 times the width of the push tow during the overtaking manoeuvre. Duringthe relatively short measuring period this value was not exeeeded. The traffie lanewidth of the tow barge was assumed to be identieal to the navigation lane width(see figure 20). The data obtained from a number of similar passing tests in othereanals was also available for the tow barge. This data is shown in figure 26.

It appeared that the relative eumulative frequeney distributions ean be representedin an approximation by parallel straight lines with a logarithmic distribution and anormal frequeney distribution along the eo-ordinate axes respectively. This means thatall the frequency distributions have much the same standard deviation.

With the aid of figure 26 it is now easy to determine the relationship between thetraffic lane width and the frequency with which this width is exceeded. In the canalconsidered above where the tow barge ean use the entire starboard half of the canal(traffie lane width 66.30 m), the frequency of exceedance is about 1%. This alsoindieates the probability of collisions. Allowance must, however, be made for a scaleeffect in the manner of steering the model, so that the probability of collisions in theprototype may differ from the probabiIity of collisions in the model. A detailedconsideration of the scale effects will be found in chapter 3.

For completeness, the relative cumulative frequency distributions are shown infigures 27 and 28 for the smallest distance between the tow barge and the push towand bank.

Finally, figure 29 shows the relative cumulative frequency distributions for the over­taking length and the time required by the push tow to overtake the entire barge train(tow barge drawn by motor vessel). This figure aIso shows the time which elapsesuntiI the push tow has just passed the tow barge and ean leave the port side of theeanal again in order to overtake the towing motor vessel on the starboard side of thecanal.

5.4 Recommended canal section

If the boundary conditions are not ehanged, the navigation lane widths (b) of thedifferent vessels may be interpreted as mutually independent stoehastie parameters(the seatter in the navigation lane width is obviously eaused by the helmsman). Theprobability of occurrance ofvalues equal to or greater than ~ bOo5 and ~bOo2 is appro­ximately equal to 0.5 and 0.2 respeetively, since the seatter in the navigation lane

28

Photo 16. Jn~lrllmcnt trolley with pholOgraphic apparalus on a boom used 10 detcrminc thepo~ilÎon of {he ships.

widths of ünc of lhc vessels is generally high in relmioll 10 thai of thc (WO others.

These figurcs show thai in addition (O the ni.lvigatioll lanc widlhs there Illust al50 besufflcielll room lef! 10 rcducc the probability of collisiol1s lo practical I)' Lcro.Account muSt be taken in this COllllcxion of:

a the ~cattcr in navigation lane \\idlh~ due 10 tllc conduct of thc hcll1lsman:b (he frequent)' at which a particular traflic silualion OCCUfS:

c Ihc COJ1seqllence~ of a possibJe collision (gcncrally consisting in onc of thc vesselsscraping along anolher or along lhc bank):

ct din"ercnces bel\\een practical conditions and thc model tests.

It may in fuwre be possiblc to sohe a deci:..ion problcl11 of this kind \\ith the aid ofstatistical and economie data. As this cannot be done al present. detailed discussions

ofthc interpretation ofthe model results look place \\iLh the canal managemelll \\hichhad cOl11l11issioned the study and practical experts.

Thc scatter of the tolal navigation la ne widlhs (l: b) provided the basis for thecriterion \\<hich was I1nally chosen. This scalter is characlcrized by the difierencc

29

between ~ bO. 2 and ~ bO.5' The index here indicated the frequency with which ~ b isreached or exceeded. The canal must be wide enough for the space (R O. 2) remainingin addition to ~ bO•2 to be sufficiently large for the probability of colIision to be keptto a minimum even when the navigation lane width is still greater. This may reasonablybe expected to be the case when R O. 2 = 1.5 (~ bO' 2 - }; bo.ó). With the aid of thiscriterion we can calculate for each passing manoeuvre for which adequate data isknown, how wide the canal must be to obtain the required degree of safety. Theresults of this calculation are shown in table I (rounded olf to the nearest 5 m).

Passing manoeuvre Canal width required in m Water depth in m

I) 2)----~------ ------ ---~~--

m-d-s 125 5s-d-m 110 5s-m-d 95 5

m-d-s 100 6s-d-m 85 6s-m-d 95 6m-d-sm 90 6sm-d-m 95 6m-d-m 80 6m-m-d 90 6

----~--

1) m-d-s

s-d-m

s-m-d

m-d-sm

sm-d-m

m-d-m

m-m-d

2)

a barge train is overtaken by a push tow with an oncoming motor vessel; respectivespeeds 7, 10and 15kph.a motor vessel is overtaken by a push tow with an oncoming barge train; respectivespeeds 8, 11 and 7 kph.a push tow is overtaken bij a motor vessel with an oncoming barge train; respectivespeeds 10, 17 and 7 kph.a tow barge, coupled alongside a motor vessel, is overtaken by a push tow with anoncoming motor vessel; respective speeds 8, 11 and 15 kph.a motor vessel is overtaken by a push tow with an oncoming tow barge coupledalongside a motor vessel; respective speeds 8, 11 and 8 kph.a motor vessel is overtaken by a push tow with an oncoming motor vessel; respectivespeeds 8,11 and 15 kph.a push tow is overtaken by a motor vessel with an oncoming motor vessel; respectivespeeds 10, 17 and 15 kph.The canal width should be measured at the keel plane of the ships.

Table 1. Canal width required to enable the above passing manoeuvres to be completed withmaximum safety.

The minimum initial navigation speeds at which the motor vessel can stil1just overtakethe push tow were determined for three initial push tow navigation speeds (9, 10 and

30

11 kph), and in two different canal sections with widths and depths of 100 x 6 m2

and 130 x 5 m2 respectively.At these minimum navigation speeds (see table 11) the overtaking manoeuvre takes avery long time. If the manoeuvre is to be completed smoothly the speed of the motorvessel must be about 1 kph higher. The results show that there is practicaUy nodifference in regard to the minimum speed of travel necessary for overtaking in thetwo canal sections.

Push tow speed in kph Minimum speed of motor vessel in kph

Canal profile

9.010.011.0

100 x 6 m 2

13.815.517.5

130x 5 m 2

13.415.317.1

Table 11. Minimum speeds of a motor vessel, necessary to overtake a push tow as a functionof the canal dimensions.

31

6 Special studies

6.1 Effects on the canal section

The banks of a canal in the vicinity of the water level are damaged by ship waves,caused primarily by fast (empty) motor vessels. Push tows do not create any extra

difficulties in this respect.

Damage to the canal bed and slopes below the water level are due to the speed of thereversc flow and propeller race, and also to the water-level drop which may causeground water over-prcssure which forces olf the slope lining or soil particles. Thesephenomena are Jikely to occur more extensively and for longer periods when pushtows use the canal than in the case of conventional inland waterway vessels.

A brief study was made of these phenomena; the water-level drop and flow velocityaround a push tow were measured, and information was gained with the aid ofmovable bank material on the mechanism of erosion. Tt was found that the erodingaction on underwater slopes is greatest next to the push tow, level with the bilge. Thepropeller race of a moving push tow does not cause any special difficulties. Figure 30gives an impression of the water-level drop.

6.2 Side-wind

An uniaden push tow in 2 X 2 formation is difficult to steer in the presence of side­wind. On the Waal, uniaden push tows therefore sail in swallow-tail formation(although with the introduction of more powerful pusher tugs, this formation hasonce again been largely abandoned). However, this is not a practical solution forcanals, because the width of the tow is limited by the dimensions of the locks to twobarge widths. lncidentally a solution may be found by providing a steering device atthe front of the push tow (bow rudder, bow propeller etc.). These aids to manoeu­vrability are being used on an increasing scale.

Another simple solution is provided by ballasting. A number of model tests weretherefore carried out in which the load state and wind speed were varied. The driftangle needed by the push tow in all these instanees to hold its course was used as aparameter to compare the difficulties likely to be encountered in any given situation.Good results were obtained when only one of the front barges was ballasted, as was

32

Photo 17. O\enaking manoeuvre of a push {wO and 10W barge lowed bya motor ~hip in IhcEendracht bend of Ihc Scheldt·Rhinc Cunal.

33

Plloto 18. Instrlllllcnllrol1c)' llscd 10 pholograph thc posilÎons oflhc ships.

already kno\\ n from practical experience. h was <1lso found (hm good results are

possible "hen lhe barges are fitlcd wilh a leeboard (see figure 31). The sludy did nolenable the maximum \\ind speeds lO be defined al which an empty push {Q\\ can stillnavigatc on thc canal: the grealcsl diHkulties oecur when the push 10\\ has to stop

or suil a\\ay from lhe lee share. bul the study \\as confincd 10 push tO\\5 n:.t\igating

at cruising speed only.

6.3 Schcldt-Rhine Canal Eendracht bond

Follo\\ing the manoeuHing study lhe queslion arises as to \\hether safety is as great

3~

in a canal bend as in a straight canal of the same dimcnsions. lllld also whether thewatcr·current velocity has aoy infiuence here. These problcms arc relevant to the

design orthe Scheldt·Rhine Canal with the belld at Eendracht in which tidal currellts

lIill probably still be encollntered untilthc Eastcrn Scheldt lIorks \ViII bc completed.Thc study showcd lhat lhe current velocity has no nOliceable innuence on the behaviour

of the ships in relation to thc water. Safety was found to be just as high. In the bendtoa (with a radius of 3.(X)() m) the behaviour scarcely differed rrom that in a srraight

canal. Thc push tow simply deviated ralher more from the required course lhan in astraight canal. In the light of this observation. it is rccommended that in callal bcnds

\\ith a radius of about 3.000 111. the width of the cross-scction should be 5 m largerthan in the adjacent straight canal scctions (sec photos 17 and I ).

6.4 Amsterdam-Rhine Canal Dcmka bend and Goyer bridge

Thc \\idlh of the Amsterdam-Rhinc Canal bet ween Utrecht and Maarssen is limited

to about 95 111. In addition there is a bcnd at this point with a radius of only 1.000 rn.lJec3usc it wDuld be very expensive to widen the c<'lI1a\ here. the canal width was an

I)hol<) 19. Dcmka bend - existing silualion.An cmpt~ motor \cssel and a push lO\\ \\Îlh onc bargc meel a laden motor ship.

35

Photo 20. Dcmb bend - widened sitlJation.Two push I{)\\~ \\ith four bargcs each meet.

invariable bOLJndary condilion for a study in which model tesIs arc LJsed lO delerminethe passing manoeuvres whieh can be completed safcly. Bccausc of the short line orvisibility (Iocally a~ low as 600 !TI) it is important for lwO push tO\\'S to be ablc to meeteach other. This appeared possiblc. A molor vessel call still overtake a barge trainwilh an oncol11illg push low. Rcstriclions fOT navigation need thcrcfoTC simplyprohibit overtaking of or by a push 10\\ immediately berorc or in the bend (sec photos19 and 20). A similar reslriction Illust also be imposed Icmporarily al other localnarrow strclchc~ of the canal. e.g. near bridgcs "ith a relatively small span. For anexisting bridge (Goyer bridge) and its sub-structure. it was determincd which trafficsituations "ere not permissiblc al this point" ith a nlll11ber of alternative canal shapes.It was found necessary to widcn the canaillp to the bridge piers (passage "idth 68 111).There is then no problem if a push lQW meets a convcnlional vesse!. T\\o push lOwscannol. ho\\'cver. pass in OPpOSilC direction at this point (sec pholOs 21 and 22).

6.5 8cunin~cl1 out let slrucfure

In conjunclion "ith the widening of the Betuwe rcach of the Amsterdam-Rhinc Canal.lhc aid pumping-station that used la discharge water from the river Lingc illlO the

36

Photo 21. Goycr bridge cxisling situation.A laden lUw meets an cmply tow.

Pholo 22. Goycr bridge - exisling situation.A push tow meets a barge Ww.

37

canal must be demolished and replaced by a new pumping-station/discharge structure.The shape ofthe flow distribution structure on the oudet side must be such that shipssailing at a short distance from it are not adversely affected by the cross-current.Figure 32 shows the flow pattern encountered with a delivery from the structure of30 m3/s, and the practically negligible influence on a push tow sailing alongside whichin the model was ultimately achieved in the recommended layout.

6.6 Maarssen intake structure

To meet water supply requirements for the central and west Netherlands, it is assumedthat in the vicinity of Maarssen water will be drawn olf the Amsterdam-Rhine Canalat a maximum rate of approx. 60 m3/s. Ships sailing close to the intake structuremust not be adversely affected; this places high requirements on the design of thestructure. The study showed that especially relatively small motor cargo vessels(length up to 60 m) are influenced by the cross-current created here. By splitting theintake structure into three rather smaller units at distances of 60 m from each other,these vessels can sail past practically without hindrance. Figure 33 shows the shipmovements as ohserved in the model.

38

Figures

NORTH 5EA

~ii~~~~~~~~~=SCHELDT-RHINE CANAL('"'") /) 0'". ._0 \'-0)' •

\ EindhovenEASTERN SCHELDT

Scale 0 25 50 75 100 km~I===5';;;;;;;;;;;;;;;;;;;;;;;;;t'===5';;;;;;;;;;;;;;;;;;;;;;;;;j1

Groninge-n

•(

I/

II

,,-.)

?'-- ._.,

)

,),-"<:. ...,

...-./

Figure 1. Inland navigation canals in the Netherlands suitable for push tow navigation.

40

.J 25

"f-of-

a.o

-

...,

~~

~

w

""f-ZWu0:wa.V>

"V>a.:x:V>

a.o0:W

'"~::>z

20

15

10

oo 500 1000 1500 2000 2500

Figure 2.

------. LOAD CAPACITY IN TON5

Relative frequency distribution of laad capacity of laden ships.

.J

"f-of-

a.ow

""f-ZWu0:wa.

V>

"

~o

".Ja.o0:W

'"~::>Z

o

0

'"L5

-

.....0

.....

- ...,5

L

~r J IL--t-...10 15 20 25

Figure 3.--+ TRAVEL SPEED IN kph

Relative frequency distribution of travel speed of laden ships.

41

Z

!w><: 100 0«l-a:W

~oZVJ 80 20 OW0.. '":I: «:I: I-VJ Ua:-wZ :I: >w ~O

0 60 40« VJI-...J 0..0LL iz

VJ0 wZa:w 40 60 w«

Cl00« «zI-

Z ...J«w LLU owa:

'"w 20 80 w«0.. Cl I-«a:

1I- Wz>wOUl-o 100 505 10 15 20 o..z

.. TRAVEL SPEED IN kph

• NUMBER OF LADEN SHIPS WHICH ARE OVERTAKEN , EXPRESSED AS APERCENTAGE OF THE TOTAL NUMBER OF LADEN SHIPS WITH THESAME TRAVEL SPEED

--<>-- NUMBER OF LADEN SHIPS WHICH DO NOT OVERTAKE AND ARE NOTOVERTAKEN, EXPRESSED AS A PERCENTAGE OF TOTAL NUMBEROF LADEN SHIPS WITH THE SAME TRAVEL SPEED

Figure 4. Influence of travel speed on overtaking.

42

'/w NI 033dS l3"''V'~.l "'W NI 31~NV l..:.ll~a--- ...-~-."' 0 "' 0 '" 0 "'0 "' ~ '" q '" q " "' '" '" '" ".; ..,

'" '" ~ I I I + + +

s/'iZlZlJ6ap NI

NOJ.l:::>31.:.l30 è13aanèl

•.. --+---

o '" 0~ ~ '"

H1GlM 3N'V'1 NOt.l'9'91J\\fN

.---HiOI'" dlHS

o0,'<f

c'Eo~

o2·

0

~

:.u....Q)

.:.:0 '"g E

Q)Q)

:r:col>

.Cl0U<07'

0 i:i~ Q),...

c:Q)....::l'"olQ)

EQ)

0.>0

0 0q 0'<f

0':;

.----,-----~

Ç/S:llllJ6~p NI

531.11:>0'3/\ ~f'iln9NV

<I ~ l"'Il C\l V .()000 000ÓÓÓOodó

, /~ , Ir',

\ 0' I

! } -.- I:,,,' ,

i l.l :" \\

1/ , ... ,- i<. . -.. -.... .'

I.....r".,..I-"" \,

", • 1""'" \I ..' , ,. ,. ,

! I,

~~ i, . -.J I,1:" "". II ) .,' '. ,- I.'

I " ..~, ;:;'; ". l~

,i ) ", .' I

L.-",

\," """"",

( .... It:: I

I '"""~

I I' '" .; ,i r..... r-. . Ic: .. ~.' .. ' I"- ,.',

I10' .' I '...,

I"-~. ~\

(,,

,~,

li "~

,I

. ~ ~ ~~.,

,~I,

I ( '1" ,, "r- I.I..... I

! "' l' '" ~,

<ri r> .. ' 1-" .l- ..- ,.., .-, ,- .. L!".,L...."" .. " r- \, '" ", " ,

! \ ".:~ ~-

jfJ' ,\

( ) ..' )" " ~l"-

~ '. .-~: ,,<

'. ,,' II'.

i -.. : ~:"" ij

r.... ~. ,I- \....

I '" ....'II " ,.... ,~

,I' .•.. 1- ,

•I ~ '\ 0) ., l"- ..1"-: \i r

.-" r' -',."

"' "'

'S'd ..._---­

'g's .------­wdJ NI 033dS

_ ll3T13dOlld

W NI )(N\f9Oèl'9'OatlV.LS 0.1 M09 .::IQ

3èllN3::) ViOèi.:.l 3')NV1510

OOOVlOOOV Mt"lC\lNT""OO

'" ~

43

DIMENSiONS IN m INARROW PUSH TOW I..... 1 z F-r-

SIDE VIEW

PLAN VIEW

PUSHER TUG "VULCAAN I" WITH 2x 2 EI BARGES

WIDE PUSH TOW

1o.00I f: :1: I ~~1---'-3--,B._00---l.otll...~~!::.-=:_=-=-=:_=-=-=-1=9--,2~1:""":=0~'t;~~6~.~5_0~=:,=:,=:,=:,=:,=:,=:,=:,=:,=:,~!.J1:I....L..------'~r PLAN VIEW

PUSHER TUG"VULCAANI"WITH 2x2 ElIBARGES

I BARGE TRAIN

i , PSIDE VIEW

:1. ~.E60.00 80.00PLAN VIEW

~---~~---------c= 99.9 0

RHINE- HERNE CANAL SHIP "OLIVIER"WITH TOWING BARGE "DONALD"

---I':C,====~~zkSIDE VIEW f

rc ~1+ 80.00 ~

PLAN VIEW

~~- 62.00 ::rTPLAN VIEW

I SELF-PROPELLED SHIPS I-51K~--=~===1.J=L2.50

SIDE VIEW r--

COASTER" MARCON I" RHINE-HERNE CANAL SHIP"ADRIAAN"

Figure 6. Principal dimensions of different types of ship.

44

I Ir

:~2.S0m 4.20m

~,"mm",m';'==4100m•

PUSHER TUG-' .

.. '

-....

. .. ------------PUSH TOW BARGES1--------

,•••••• 1.0_.-';..-.-- - --- ---- ---- ---- - --- ---- - --- ---- - --- - --- -

.

e 10u

~

Cl. 00a:0 10...JW> 20w...JI

a: 30wl-oC(3: 40

.j:>.VI

Figure 7.

II__...~ SURFACE'

[l]rmtJ. BOTTOM !______1_-Flow pattern and water-level drop around a push tow in a canal.

w

,

II

DIRECTION OFTRAVEL

DIRECTION OFTRAVEL

v

~,

DIRECTION OFTRAVEL

DIRECTIONTRAVEL

W = Resistance force

v Propulsion force

D = Drift force

R = Rudder force

N Canal bank force

Figure 8. Influence on drift angle of point at which canal bank force acts.

46

15

···~~

t-ZU)

~a::>

10'""U)

:>a:U)0-

U)

--'

"Z"a:U)

00:>a:

Z

"U)

:>

1

~

WEIGHTED MEAN

~------

~~SCATTER IN MOO~'////SCALE 1:25///:;

oo 10 20

Figure9.

----. DURATION OF MEASUREMENTS IN min.

Mean rudder angle and duration of some prototype measurements with width of100 m.

· + 0.5·~·~~

t-ZU)

~a::>

I~'""U)

:>a: 0U) 0 200- IN min.U)

--'

"Z WEIGHTED MEAN

" --------t SCALE 1: 25

"0z

"U)-0.5:>

-1.0 L.- ---.L --'

Figure 10. Mean drift angle and duration of some prototype measurements with width of 100 m.

47

..Jc(

Zc(

ulilo

.. a:

2 :o~ ~

i!lilZ >lc( Za: c(~ c1l

..Jc(

Zc(

u

::Eoa:u..

10.---~---------,.-------~-------~-------~---'

5 1--+----------t--------+-------+--------,--+--1

RUOOER

TRANSVERSE FORCEBOW = TRANSVERSEfORCE STERN

ol-+----.....I,1I--+-----,....~.....tIliF,....!::....---+_-----_+__1

5

10 L--L....,- ---L-::- ---I.-_---I.-L..:---L ...l--::- --l..---.J

_0.50

00 + 1.0

0+ 1.5

0

DRIFT ANGLE .. STERNFROM

OF SHIP SLEWED AWAYCANAL BANK

MOOEL SCALE 1: 10.5

_ _ MODEL SCALE 1: 25

_. _ MODEL SCALE 1: 40

THE TRANSVERSE FORCES EXERTEO ON THE PUSH TOWARE ZERO IN THE FOLLOWING CONDITIONS :

MODEL SCALE 1: 10.S RUODER ANGLE + 10.0°DRIFT ANGLE + 0.61 0

1 : 25 RUDOER ANGLE + 10.0°DRIFT ANGLE + 0.46 0

1 : 40 RUOOER ANGLE + 9.4 °DRIFT ANGLE + 0.32°

Figure 11. Transverse force bow equals transverse force stern.

48

+ 20

" +10Ol0-

~2e0-~..c2

I~UI 0U0::0U-

UIlil0::UI>lilz«0:: -10f-

-20

, Iir+ KV

/'

W #KA

CORRECTED KA-L1NE FOR 0( -40KA+-

,~ .#~'

A

[,.~ V~

~ V~--...:~~~-'-=~

? ~=-,

.~~ :--,

~}v~CORRECTED Kv-lINE FOR CX= 25

-~~V

-0V I0 ~ =10'1ty X 0<. = 2S

• C( = 40

KA

-"V

50 40 30 20 10 o 10 20 30 40 50

RUDDER ANGLE TO PORT ....f--+---l.. RUDDER ANGLE TO 5TARBOARDIN d~9r~~s IN d~9r~~s

Figure 12. Initial transverse' orees as a eonsequenee of rudder deflection.

49

~

z<dl

...J<z<U

20VI

18

0

lAOl

16~

'"Ol.."

Z- 14w...JL'.lZ< 12cr:w00~ 10cr:

~~

cr: 8dl

...J~

06w

14

2

SHIP: PUSHER TUG "VULCAAN" WITHwater depth: 2 x 2 NARROW BARGES ~

_ •••_... T 21 3.80 m DRAUGHT OF BARGES 3.20 mT 1 4.20 m

_._._._ T2 4.50m__ • __._ T 3 5.00 m___ T4 5.20m__.__ T5 5.80 m__________• T 6 6.00 m

-_•••_- T7 6.50m-_ •• _- T22 7.00m V -_ ••••• _ T23 7.40m /' .//'. A~-:::::;

/','it.:$.~. ~ ~

:~/,/".~...-.- ~I i' .~./ ~

I----t------\----+----+-------,v.,' -::"'1/.'- ~ '/

~#J'~·r.~~.#V ~~ .-,. .. / /..' ~.._--- -,' ~/' /' I,

/_.1-' " ..,~~v/ ~::.V.; .#~';~JI""""-?

~. ,; ~."""7' .. - -- /." ....:'~::p V/ ~'~~:'p~." ;.9 .

~- ..:~ ~",.-;,',~" "/

':~~~~~I' L.--'"......y4i"~...:..-....----oo 5 10 15 20 25 30 35 40 45 50

---I". DI5TANCE FROM CENTRE OF BOW TO CANAL AXI5 IN m PROTOTYPE

Figure 13. RUDDER ANGLE. fnfluence of depth (regtangular cross-section).

U

/.,.- '"

.;",

SHIP: PUSHER TUG "VULCAAN" WITH "..,/2 x 2 NARROW BARGES ;DRAUGHT OF BARGES 3.20 m /

.Y.....

~/ ..,.....-. 1'- ......

V .,." ......

/",....

'.,.' ---.....~,.. ---' '.,.' =~ -....... .... \~~::: 1"'-.. "•• ,0::: ...--' -' .- ---

""""=1"---" -.. r--, '.-.......-...•.:.~

" . ......... ", \ \. '. .....,

'..", ""..

\ \\"" ' .. " ".. \.'. .'\. '. .. ,

'\'" '\ .. .\..

watqr Mpth: '\ \~\

-"'-'" T21 3.80 m ........ ", .- T1 4.20 m \ ..:\ "\\_._.-.- T2 4.50 m-._.- T3 5.00 m .

- --- T4 5.20 m .--.-- T5 5.80 m

\ ...\ \-----_.. --_. T6 6.00 m-_ ...-- T7 6.50m- -_ ..-- T22 7.00m

\ \-_.... - T23 7.40m

+ 1.2

'"+ 1.0

z<{lil

...J + 0.8<{

Z<{

U

\!)+0.6

z" ij..~

<{

'""-.. + 0.4

." ~

~0lil

n+0.2

0

:::;: -0.2:::>ë( Vl

lil X:::; <{

:; ...J -0.4

0 <{

W Z<{

U

\!)-0.6

Zij<{"- -0.8~0

'" -1.0

_1.2o 5 10 15 20 2S 30 3S 40 45 50

VI- -. DISTANCE FROM CENTRE OF BOW TO CANAL AXI5 IN m PROTOTYPE

Figure 14. DRIFT ANGLE. Influence of depth (rectangular cross-section).

l,L.~V ~~b.- I ~,:,.-= . .

i-X1--

\ [J ~,/ c;;

-R~•

tlr-r- r r· ~. ...". '--; '1.--+t- r

.....~ I1.\- \. --'-:"~ - ~

tJ V '- J " \J 'ol 'f 1 ,, -,

UI BB

i!~

Ir SB

o

2 4 6

TIME IN

8

minutqs

10 12

BB

HSB

PROTOTYPE RESULTS (Sqq fig. 5)

+ 2 0 ,--r--r---,---,----,----,-----r--.---.----,---.--...........,

o

_. - -- '7' .--"7"," - .-_.-_.-_. -_.~~~ ~ -::;0

1'- '- '-~ ...~

UI BB

i!Ir SB

o

2 4 6 8 10 12

___ MEAN VALUE

---I.~ TIME IN minutn

MODEL RESULTS

Figure 15. Behaviour of prototype push tow and model push tow with automatic steering.

52

o"

xI

J

)Xl.',,I

\ \~'\ /

\ /\ /-'-,' /

~-=.:.-==..-=~ /;, ~,'X

" ~ ~·~X~·/ \"/ "

/~---. -

0 '10 2b· 30 40 5~ DISTANCE ~M_ CANAL AXIS IN m

\.t.

\

X

'12 '1.2

'10 '1.0

- 2 -0,2

8 0.8lilC» lilC» lOlL-Ol C»

"L-Ol"0

"6 "0 0.6zz

UJ...J UJC> ..JZ C>« Z4 « 0.4exUJ t-aa u..-::J exex a

t2

t0.2

0 0

- 4 - 0,4

-----

Mean rudder angle

Standard deviation of rudder angle

Mean drift angle

Standard deviation of drift angle

Draught 3.20 m

Propeller speed 300 rpm

Rectangular canal section;

width 100 m; depth 6 m

Figure 16. Dynamic equilibrium of rudder angle and drift angle as a function of distance betweenpush tow and canal axis.

53

Minutes of the Netherlands Water Police dated March 22, 1966At approximately 13.30 on 21 January 1966 the following accident occurred close tokm 5 on the Amsterdam-Rhine Canal, involving:

A motor ship 'Rudolf Tiedtke'load capacity 1168 tons, carrying 662 tons of wheatdimensions: 73 x 8.24 x 1.92 mengine: 600 hp

B motor ship 'Dirk Jan Boon'load capacity 745 tons, carrying 630 tons iron scrapdimensions: 60 x 7.24 x 2.10 mengine: 630 hp

C tow barge 'Boezemsingel'load capacity: 1399 tons, carrying 1360 tons of oredimensions: 80 x 9.502 x2.53 m

The initial speed of the ships was 8 to 10kph.

After an unsuccessful attempt, the motor ship 'Rudolf Tiedtke' was again overtakingthe 'Dirk Jan Boon/Boezemsingel' tow. The suction effect caused the tow barge'Boezemsingel' to veer over instead of remaining straight. At a given moment the'Boezemsingel' veered over so far to port that it collided with the 'Rudolf Tiedtke'although the latter was sailing well over to port. The 'Rudolf Tiedtke' was damagedin the collision.

The skipper of the tow barge 'Boezemsingel' was sailing right over to starboardbefore or at the beginning of the overtaking manoeuvre; this considerably increasedthe suction effect of the bank. In addition all the ships were sailing too fast duringthe overtaking manoeuvre. Subsequently the overtaking operation was completedwithout difficulty at very low speed. lf the canal had been wider and/or deeper at thispoint the suction effect would have been lower.

Amsterdam

AmstC2rdam - RhinC2 canal

A. Motor ship 'Rudolf Tiedtke' B. Motor ship 'Dirk Jan Boon' C. Tow barge 'Boezemsingel'

Figure 17. Collision 4.

54

VlVl

0 15 30 45 60 75 90 105 120 1351150116511801195 210 225 240 255 270 285 300 315 330

• TIME IN s

, ,I, ,

,'I

tI

l

'I, I,,,,, td~~,,,, --lO~,,,, B-A = NAVIGATION LANE WIDTH. REARTOW BARGE

A+C = SPACE NEXT TO NAVIGATIONLANE OF TOW BARGE

RECTANGULAR CANAL SECTION

WIYH: t°j' DErH

:,6m IFigure 18. Picture of ship positions (recorded photographically).

BB./ I\..

iE '"J' ""' "" ...,.SB .... ...,.

300

300

300

120 180 240

---i•• TIME IN 5

120 180 240

---i.. TIME IN 5

120 180 240

---:'I". TIME IN 5

60

60

60

o

o

N

~lilCOlCOl...C7'COl't:I

lil

Zl-0

T"

:~ )(

~lilCOlCOl...C7'COl't:Il'l

I

~OT"

.~ )(

'"'z ~-..., 8Ul N...JIC) 0 4zT'"« )(

lil 0.~ COl

U. ~a::r 4O't:l 0

UI

i~~~~~ 0 60 120 180 240 300---:i.. TIME IN 5

BB/ "

,/ "- ........,/ '-/ , L

\. LSB ~

z0 4~0:: 2UI...Jl'lUIl 0~O«T'"

2...J )(« lil0::'UI E~

«z...J_

o 60 120 180 240---:I.. TIME IN 5

300

RECTANGULAR CROSS SECTION;WIDTH 100 m ; DEPTH 6m

56

300120 180 240

--~~~ TIME IN 5

60

BB 11" """""- .-J r-' ", 1\

J ~

I '\.I '"I -

SB If......,8

o

8

4

4

o

zI/)oUJUJa..I/)N

I

.JO0(T'"a: xUJf- UI0( .......J E

~100

0 80a:~u. zEo( 60al

~o 40UJa:uo(zO 20o(alf-a:

01/)0(-f-Ol/)

0 60 120 180 240 300

~ TIME IN s

:-1 'fd f EB1.5

~E

Z

oUJUJa..I/)

o 60 120 180 240

--~.,. TIME IN s

300

1\I \

I \ "A \ j \... f~ T7 ~, \ IJ V "'\ 11-,

z + 1.2

z +0.80~N

+ 0.40(1a:0UJT'"

0.J XUJ NU UI-0.4U ......

0( Eo 60 120 180 240

---I~" TIME IN s

300

Figure 19. Results derived from positions of rearmost tow barge.

57

50

60

~o~

u­o

•

•

•

•

• • •• • •

• • • • •. • . • •• • • • •• •• • • • •• • • •

• .-

5040302010

30

10o

20

40

f

wz<...J

zo~\.?;;<z

-------. TEST NUMBER IN CHRONOLOGICAL ORDER

Figure 20. Passing manoeuvre in which tow barge is overtaken by push tow.

•

••• - .

• •• •• •. •• • • • •• •• •• • •

• • • ••• • •• • • •

••• ••

w

~i 25

~~~ ;0o 0~~

u-o0 .. 20

c:w "Oei.Vi_

Ol

~~

~ ti 15dl-oa: ~< ::>~ ..lil "Ol

,,~

~ E 10U-z

w'"Uzz<<dl~

~ 0 5O~

r oo 10 20 30 40 50

--+ TEST NUMBER IN CHRONOLOGICAL ORDER

Figure 21. Passing manoeuvre in which tow barge is overtaken by push tow.

58

•• •

•I. • .. • • ••• • • •

• .• • •

• • •• • ••

• • •• •• • •

• •

0 10 20 30 40 50

-----+ TEST NUMBER IN CHRONOLOGICAL ORDER

•10

Passing manoeuvre in which tow barge is overtaken by push tow.

50

0~

w 40l.:>a:«l1l

~E

Oz~- 30LL~00~

weIVlVl

::> 20~a.a:OLL0.0

:::!:woea: Vl

10LL0

wa:u«zO«11l~a:Vl«-~ 0OVl

1Figure 22.

O~ 90~ Ol

'"~ :;O.D~ ~

I 2Vl ë 80::>a.

Olc:

LL "Ol>.

w-..e~Vl

o "cc",..: Ol

o ~al ..cc g..: E~~Vl

E

tl~z«'"~z

~«011l

70

60

40

• . .• •• • .

•• • . .• • • • canal axis

f--- -- f--. ~. ---L ....... f'- - -0--- - -+--.. •• • • •• • •• •

•

r 30o 10 20 30 40 50

----. TEST NUMBER IN CHRONOLOGICAL ORDER

Figure 23. Passing manoeuvre in which tow barge is overtaken by push tow.

59

60

Ib 40

~

3:oI-

u­o

e

~

w 50\:>0::<{à:l

."§j-,

~I• .1

I

I,t,

• !I!

.1 •

1

i. • •I • •

• • .. :j : ••. o.••• r.. • •

• .-.- : •• •• I·

20

30

t

wz<{...J

zo~\:>

~Z

50 60 70 80

---. DISTANCE STARBOARD SI DE OF PUSH TOW TOBANK IN m (m2asur2d at tow barg2 k221 plan2)

Figure 24. Passing manoeuvre in which tow barge is overtaken by push tow.

60

o ez<~

I-~I/) Z 20UI<...J1O...J<0~z1/)<

oo

A

f- canal axis -- -

•I-

•Ie"

• • I ..

• ~ IA .- A,,-• • ••.-.

IW' Itl •

-I"

0

ll' ""r'..o.0 klc 1.- .... 0

!jO "o ~ o CO

" I'" P 0

0

UI~et:<10

~,...

0111I-f'

aZAUI ~UI 0~ ..1- ....UI 010

lilI/) c:UI~u Cl.Z_

~ =1/)-"

oC;I-"'CI'I/) lilUI ..~ ~et: a< lil...J e~

f

80

70

60

50

40

30

10

10 20 30 40 50 60 70

---I~. NAVIGATION LANE WIDTH OF TOW BARGE IN m

•o

LARGE5T

SMALLEST

DISTANCE

DI5TANCE

FROM BANK

FROM BANK

Figure 25. Relationship between navigation lane width of tow barge and its distance from bank,

61

I ! IW').o,...CDCDtn..,..,,&nln

1 · r-r-r-r-r-I I ~~~~~CD '" ",I-- ,..,

lil '" .. ,..,• z r- .. Ul~• 0 l?

I I ï=,..

'" ",I-- a: ~ ~ ~O

'" 5! • -< ...... ~u r- IOlil :rlil '" '" I--

~ ..... U')i '" 0 '" 0 ...... :::)

...J r- .. r- a.

! -< 0IJ IJ

'"., z '" '" '"J:

-< r- .. r-a ........• • u ........• L; J: i• J:...J b: 11 -< r- Ul

a i :l;- z ........l- IL, '1 ö -< ........

~,La ..J

N J;Ul 0 ~I-- za.

\. 1ï.~-\~.., -< r- 0

a: r- i;( 1=0

"\r- IO ~ -< .........

l? ........\ \~, ~ il

~z

~rL '''' \ I·III" ·•ot ·I i I'11 \ , ..... ·•·.., "\ .... ..., ~

h....1., " I" 1,,,-\.

lil I. • to, \.·1_.., 1 :..,

"'t .. .:; il, "':

I

L,I L I ~-r '1

•1 ..I

II•I

!I

H.LOIM 3N'tl NOI.L't~IA'tN

~O'O

50'0

Z'O

~

5

! Ol-

UIUz OE<0UIUIU 05)(

UI

U.0 OL

UIC)

~z06UI

Ua:UI 56a..

66

9'66

56'66

66'66

N

CD .0

T"" Ó Ó

H.LOIM dlHS ...N

Ó.-;o

66'66

56'66

9'66

66

56UIUz<0UI

06 UIU)(

UII

zOL

0zu.0

05UIC)

~OE z

UIUa:UIa..

Ol-

f5

~

Z'O

50'0

~O'O

Figure 26. Passing manoeuvre in which tow barge is overtaken by push tow,

62

1100

r ~LIJ rU-zeo

f«0LIJLIJ

t--- u

1,1)(

LIJI

Z60t--- 0

Z

u.t-- 0

LIJ (l!l«t--- I- 40

Z

~LIJUa:

t--- LIJ11.

/t--j 20

)I ~

0-10 o +10 +20 +30 +40

---+ DISTANCE BETWEEN NAVIGATION LANES OF TOWBARGE AND PUSH TOW IN m

Figure 27. Passing manoeuvre in which tow barge is overtaken by push tow.

V I----'

(11

1/

JJJ

r.....U

~Vr-

100

wUZ eo«0wwU)(

WI

Z0 60z

1)wl!l«I- 40ZwUa:w11.

j 20

oo S 10 15 20

~ SMALLEST DISTANCE BETWEEN TOW BARGEAND BANK IN m (m..asur..d at k.... 1 plan.. )

Figure 28. Passing manoeuvre in which tow barge is overtaken by push tow. 63

201B16141210B6

--~ --- _r- I _. _. .. ...• .....J

r0-

I-:"'-' r.

1" i_.r-" !.-J__ I

......p.

~I- •.J .....i i ,.....rI .-•

J- f~. ..... J.. li..I

/I ,..- ~I r --

~rEI'

I• .J"

p- ~-j ro

20

Ba

40

60

f

tJ 100Z«oUIUIUxUII

ZoZ

U­aUICl

~ZUIUlI:UIQ.

---i~~ OVERTAKING TIME IN minutu

1.6 1.B 2.0 2.2 2.4 2.6 2.B 3.0 3.2

.,............ OVERTAKING DISTANCE FOR ENTIRE TOW IN km

TIME NECESSARY Ta OVERTAKE TOW BARGE

TIME NECESSARY Ta OVERTAKE ENTIRE TOW

TOTAL OVERTAKING DISTANCE

Figure 29. Passing manoeuvre in which tow barge is overtaken by push tow.

64

Lines of equalwater-level dropin cm prototype

Wide push tow;rectangular ca­nal section;width: 100 m;depth: 4.20 m

--10.....~-~·<O ...······30~ '

......·-40·.. ···

"-,."\

( \

: I\. 0\..,1rI

(',,. '

I \~ \I ~

" \b

O"-

~ \I X)t I\j

" .,,0­'. I'-.,

- -'-'40_

..···...... '30....

120..\' ..-

,'-

~o

sa_

..'.. '

·•····· ..... 30 ...'. .... ...... ..........

'-'-',\,\IiI,i

I-'-40''''

~I.Ul_-10_ 1: 1)(1-- __ ...... " ..........., .,- "~t,,,.,., '-

---20 " I,~---------_ - ~ ,-,. ~ ..'.' ,-

.....···..·······30.·...·... .)/,............. I

I ',I

/i

I/

/iI;I\i,\"40_, I,

",(-' -

-----20_...

............'70,

'-v = 10.9 kph

Figure 30. Water-level drop around moving push tow.

65

9874

DRAUGHT OF ALLBARGES 0.75 m

511-+-~--+-.......jj~--1----7I'-:.......,I,,~--+-4---.----l

o _"""'-_......_ ......_-:!:::,...I111111::;;;,j".__......__.....3.•3.,;0.;,;m;...e_----1

01 2 3

1 5 1I-+-~r---+-__+---1------:li:---+--I---+---lz

f

UJ...JC)

~ 1 0 1I-+--t---1~-+---iII~--+--I--+------,j~-----l

~u.a:o

---\I". WIND STRENGTH ON BEAUFORT SCALE

20

//

/

15105

I I I IDRAUGHT OF OTHER IBARGES 0.75 m I

511-+-~r__-+-__+---1'+_:~-~'---"""'*'~--+---l

z 1 5 1I-+--t----1--+---+--+--I---I-t----+-----/

f

~u.cro

20 It-+--;--+---+--t---+---+----+---~= LOCATION OF BALLASTED BARGE:

~ WINDWARD SIDE FORWARDg' __ LEE SIDE AFT."

UJ...JC)Z0( 10 DRAUGHT OF ONLY ---+----lI!'---,jI[J.....

BALLASTED BARGE 1.50 m I&.JL.---:--

---\I". WIND SPEED AT HEIGHT OF 10m IN mis

Figure 31. Influence of wind and draught on drift angle.

66

'Tl • 200;;' Wt:: ...J.... t''" Z...., <{ ..

!"" eI: ~ + 0W '"

tl:l o Ol

'" 0":r-

~ ~ ,III 20<ö'

4t:: W ·.......J

0 <.:l...,Z

'0 <{ ..t::'" w~+ 0:r-a lil '"eI: Ol

~ ;:>"

::l 8 ~ , 4'"III....o:l J, E · 4'" o~t:::;S' ~ ....

OQ eI: Z + 0'" WW::l >~0 lIlWt:: ZU- <{<{

~ ~ ïi , 4'"....t::()

ë....'";0II

....,0

i

RUDDER ANGLE

PORT i DRIFT ANGLE

~

- - - ----- -- -STAR BOARD

PORT I I I OFF SETTRANSVERSE DISPLACEMENT- ~--!------~--1--- _.-- ---~--~--~--~----

CORRESPONDING VALUES

STARBOARD I I400 200 I 0 200 400 600

II

I i II "-O'05~ I

I STEM P?fuION I "\ ' / IL..-- _ -+- - -----i - -- -t-~- - =++~g,-----t - --- --1--- _ -+----- - --+--- _ ---+-- _ ---+- _...J

I 0 45° .--11/4 BOTTOM WIDTH f -E-?Ç 3- --- -------~-==--1

, .~ _O.1~--- _-------- I

~'\~ _o.~~~=f8·2?==sw= .. == g.C:; WATER-LEVEL NAP. + 2.7öll\~

0\-.l

PUSH TOW

LENGTH : 191 mWIDTH : 22.80 m

DRAUGHT : 3.30m

PROPELLER : 200 rpmSH Ip's SpEED: 2.0 mis

PLAN VIEW

scal~ 1: 5,000

PLAN VIEWseo Ie 1: 5,000

PROPELLER SPEED 120 rpm ;SPEED 2.50 m/s

350

DI STANCE FROM FRONTOF MOTOR SHIP IN m

4 I ~

150 0 150

PORT II

I

I------ -- ---&..._-~.----~.----

-i'-.... r-' ~ ..-......../

100..STARBOARD ~

P6RT Iy ~ ............. ....... ----

_.! -1---- - ---=------ -

STAR~OARD I

PORT !----- - --~----- - --

~ ~

STARBOARD

3506

6

8

o

i'j 0WOllIl~CC Ol=>""OzU-

a

20WtI 0CC~W'"C Olc""=>zcc_

20

I

lil EIC~

~~CCw~~lIlWzU«CC...J~a.

Figure 33. Behaviour of motor ship near Maarssen intake structure (Q = 65 m3 /s).

68

In the series of Rijkswaterstaat Communications the fo\lowing numbers have been publishedbefare:

No. I. * Tidal Computations in Shallow WaterDr. J. J. Dronkerst and prof. dr. ir. J. C. SchänfeldReport on Hydrostatie Levelling aeross the WesterscheldeIr. A. Waalewijn

No. 2. * Computation of the Decca Pattern for the Netherlands Delta WorksIr. H. Ph. van der Schaaft and P. Vetterli, Ing. Dip!. E.T.H.

No. 3. * The Aging of Asphaltic BitumenIr. A. J. P. van der Burgh, J. P. Bouwman and G. M. A. Steffelaar

No. 4. Mud Distribution and Land Reclamation in the Eastern Wadden ShallowsDr. L. F. Kamps t

No. 5. Modern Construction of Wing-GatesIr. J. C. Ie Nobel

No. 6. A Structure Planfor the Southern IJsselmeerpoldersBoard of the Zuyder Zee Works

No. 7. The Use of Explosives for Clearing leeIr. J. van der Kley

No. S. * The Design and Construction ofthe Van Brienenoord Bridge across the River Nieuwe MaasIr. W. J. van der Ebt

No. 9. Electronic Computation of Water Levels in Rivers during High DischargesSection River Studies. Directie Bovenrivieren of Rijkswaterstaat

No. 10. * The Canalization of the Lower RhineIr. A. C. de Gaay en ir. P. Blokland

No. 11. * The Haringvliet SluicesIr. H. A. Ferguson, ir. P. Blokland and ir. drs. H. Kuiper

No. 12. The Application ofPiecewise Polynomials to Problems ofCurve and Surface ApproximationDr. Kurt Kubik

No. 13. Systems for Automatic Computation and Plotting of Position Fixing PatternsIr. H. Ph. van der Schaaft

No. 14. The Realization and Function of the Northern Basin of the Delta ProjectDeltadienst of Rijkswaterstaat

No. 15. Fysical-Engineering Model of Reinforeed Concrete Frames in CompressionIr. J. Blaauwendraad

No. 16. Navigation Locks for Push TowsIr. C. Kooman

No. 17. Pneumatic Barriers to reduce Salt Intrusion through LoeksDr. ir. G. Abraham, ir. P. van der Burgh and ir. P. de Vos

No. IS. Experiences with Mathematical Models used for Water Quality and Water QuantityProblemsIr. J. Vaagt and dr. ir. C. B. Vreugdenhil

*) out of print

69

In the series of Rijkswaterstaat Communications the following numbers have been publishedbefare (continuation):

No. 19. Sand Stabilization and Dune BuildingDr. M. J. Adriani and dr. J. H. J. Terwindt

No. 20. The Raad-Picture as a Touchstone for the Threedimensional Design of RoadsIr. J. F. Springer and ir. K. E. Huizinga (also in German)

70

-N800

~00

~ww­*