Efren Ed. C. FLORES and Kei NAKASAI On the Flow Distribution of an Experimental Circulating Water Tank

(1)

103

On the Flow Distribution of an Experimental Circulating Water Tank

Efren Ed. C. FLORES and Kei NAKASAI

The study was made in order to improve the accuracy of experimental procedures in tank experiments of model fishing nets.

The characteristics of water flow distribution of a new experimental water tank of the circulating type was studied. Water in the tank is circulated by a propeller powered by a 3 phase A. C. motor. Standard flow equalizers included 2 sets of meshed conduit plates and 4 sets of angular conduit plates. Two series of experiments were conducted ; one with 4 sets of meshed wires and the other with none. The former series showed a better water flow distribution than the latter.

Introduction

The advantages of studying the mechanical characteristics of nets in model form using water tank experiments have been expounded by T. KAWAKAMI1).

He further stated that in testing models, the law of similarity between model and full-scale nets, which was first introduced by M . TAUTI3), must be observed.

However, aside from this, the characteristic water flow distribution of the water tank to be used for the experiment must be investigated first to get a good accuracy of the model net experimental procedures.

Water tanks are generally classified into two categories according to the state of the water contained. First, the water is stationary and the model net is towed from a carriage moving on rails at designated speeds. The water tank is that of a long canal provided on the side wall with glass observation windows at fixed intervals. Such type is used in the laboratories of Tokai Regional Fisheries Research Laboratory, Tokyo ; Department of Fisheries, University of Kyoto, Maizuru ; The Nippon Gyomo Sengu Kaisha, Ltd. , Shimonoseki and others.

Second, the model net is held stationary. The water tank is of the circulating type with water circulated by a propeller or by other means. Examples of this type are those in the laboratories of Tokyo Fisheries University, Tokyo; Faculty

of Fisheries, Kagoshima University; Nagasaki University and others.

This report gives the characteristic speed flow distribution of the second type.

Y. NARASAKO and M. KANAMORI2) conducted a similar study on the performance of a circulating water tank of twin symmetric elliptical circuits. The

(2)

104 Bull， Fac． Fish．， Nagasaki Univ．， No． 32 （1971）

bilateral symmetry of the waterway produced a good natural effect as the water flows through the central waterway coming ・from both left and right waterways．

Since the tank under investigation in this report is，only a left or a right half portion of the above twin tank， the results of the performance is different as will be seen in the following discussion．

Circulating Water Tank

The circulating water tank as shown in Fig． 1 used for the proceeding model net investigation is of the rectangular type， 600cm long， 400cm wide， and waterway made of welded plates mounted on concrete base． The waterway，

80cM wide and 80cm deep， is uniform throughout the circuit・

Water is circulated by a propeller powered by a 3 phase A． C． motor． The

600 一一一一一一一一一一一．一一pt一

A誌面馳

T︒︒q一⊥

？一 r

／／

G； ^B c

432−OOOO OOOO

D

r

D

o o o o

4 3 2 1 0 0 0 0

ロX x

1 2 3 4

S 納 B

o o o o 潤@o o o 潤@o o o

lX

トー80一一一一・一一，1

va11

T−eoO一﹂←﹇馨

Vll E VI

二千

v

1 8 8

IV

o o o o

1 2 3 4

o o o o D

0000

D

OOOO OOOO OOOO 12340000

C

轟OO

Fig． 1． Top view of the experim ental tank；（A） motor，（B） propeller，

（C） meshed conduit plates，（D） angular co nduit plates and （E） glass observation window．

(3)

E． FLoREs and K． NAKAsAI ： Flow Distribution of Water Tank 105

maximum speed flow measured at the observation window is 70 cm／sec without the meshed wires and 54cm／sec with the meshed wires．

Water circulation is clockwise， starting at the propeller in the left straight waterway with the first meshed conduit plates， then to the right straight waterway through the first two corners with fixed angular conduit plates． From the right waterway， the water goes back to the propeller area through the 3rd and 4th angular conduit plates and is recirculated．

The observation window， 50cm high and 80cm wide， is provided on the inner wall of the right straight waterway where photographs can be taken of the model net being tested．

Method

The A． C． motor revolution scale is marked from O to 10 with O．5 calibra−

tions． Measurements of water flows were taken 10 minutes after changing the speed of the motor．

The water flow was measured at 11 reference points （1 to XI） around the tank as shown in Fig． 1． The water flow was examined in three dimensions ； 1） through 5 steps （V to ，IX） in the direction of length along the right straight waterway， 2） through 4 steps （sub．1to 4） in the direction of breadth and 3）

through 3 steps （S−surface water， M−mid−water， R−bottom water） in the direction of depth． The water flow was measured by an electric current meter （（tDenta CM−IB Model）．

The circulating water tank was filled with tap water 72 cm deep． During the experiment， the water temperature ranged from 25 to 26 degrees centigrade．

The first series of experiments was done using sets of meshed conduit plates

（C） in Fig． 2 and angular conduit plates （D） in Fig． 3 as flow equalizers． A set of meshed conduit plates was placed at 100 cm down stream from the propeller and also about 50 cm before the observation window （E）． The sets of angular conduit plates were placed at the curves of the tank． Measurements of water

面面

Fig． 2． Meshed conduit plates．

(4)

106 Bull． Fac． Fish．， Nagasaki Univ．， No． 32 （1971）

Fig． 3． Angular conduit plates （D） and meshed wires．

flow were made at intervals of O．5 on the motor scale from 3 to 9．5 as well as

at 9．8．

In order toe improve the uniformity ， of the water flow， a second series of experiments was carried out with meshed wires measuring 2 mm in diameter of different mesh sizes． These were placed at appointed points along the waterway，

i．e．， before the first meshed conduit plates （mesh size＝33．5mm）， first angular conduit plates （mesh size ＝26．1 mm）， second angular conduit plates （mesh size＝

22．3mm） and second meshed conduit plates （mesh size＝13．0 mm）．

Measurements were only made from reference points V to IX both for vertical and horizontal distribution of flow speed， at intervals of 1．0 on the A． C． motor scale from 3 to 9 as well as at 9．8．

Result and Discussion

Vertical cross sections of the flow speed distributions were drawn for all reference points （1 to XI）． Horizontal cross sections were drawn from points V to IX for steps S， M and B． All figures were taken at 6．5 motor scale calibra−

tion except for Fig． 5．

Reference points 1 and JI 一一 The vertical cross sections at reference points I and II show maximum speed at M3 decreasing elsewhere in all directions away from this point． This manner of decrease in flow speed produces closed isovels in somewhat concentric circles as shown in Fig． 4． However， at motor calibration 9．8， the circular pattern is broken at points M3 and B3 having almost the same flow speeds as shown in Fig． 5．

The velocity is lowest at the inner wall （Si，tMi，Bi） and gradually increases towards the outer wall with maximum at S3， M3， B3． Then the velocity again drops as the outer wall is approached． This is the general trend with all reference points． lt is safe to say that the increase towards the outer wall is caused by centrifugal force， and the velocity decrease near the walls， including also the bottom， is caused by water viscosity．

Reference points III and IV 一一 After the water passes the angular conduit

(5)

E． FLoRES and K． NAKAsAI ： Flow Distribution of Water Tank 107

plates at the first corner， the vertical cross section at reference points IIT and ZV show a different picture from that of reference points f and II as shown in Fig． 6． The maximum flow speed is located at point S3， followed by．M3 and B3． This shows that with respect to the lateral area of increased velocity，

reference points III and JV are similar with reference points 」 and ll． However，

the vertical distribution is different．

There is a marked difference between the surface water （S） and the mid−

water（M） producing diagonal isovels towards S3 and M3 coming from both sides and the difference increases proportionally with the increase of water flow．

茸 ︑ノ8

20 30 40

（9i

40 ^30 t

＠2

Fig． 4． Vertical cross section of flow speed （cm／sec） distribution at reference points 1 and II．

joo

tt

40 60 80

100 80 oo

Fig． 5． Vertical cross section of flow speed distribution at reference points 1 and II at 9．8 motor scale calibration．

The marked increase in velocity of the surface over the mid−water is because the former is free from the bottom and is only subjected to wall friction on both sides． The mid−water and bottom water （B） do not show a marked difference thus producing more or less vertical isovels，

Points SI and S20f reference point皿 show a distinct difference from that of reference point IV． The drop in velocity between point S2 and Si is bigger at

(6)

4 5

@ 磁

撃u

v

so

N

x

︶ililliZil

N

de

Fig． 6． Vertical cross section of flow speed （cm／sec） distribution at reference points III and IV．

ロV

30

xss

de

ww

rp

Fig． 7． Vertical cross section of flow speed （cm／sec） distribution at reference points V and VI．

reference point皿than at reference point JV． Water flowing to Sl of reference point 1−II from S l of reference point ll is weak． However， the water passing through Sl of reference point III picks up speed as it goes to reference point 刀7 being affected by the waters of S2 of reference point JJI which also moves to the same reference point．

Reference Point V and Vf 一一 After passing through the second set of angular conduit plates and meshed conduit plates， the flow speed distribution is very much improved． This is shown in Fig． 7 by lesser soft diagonal isovels for the areas of Si，Mi，Bi to S2，M2，B2 as compared with previous reference points．

The pattern of the isovels is similar to that of the previous figure． However，

the number has decreased 1 showing a decrease of difference in cross sectional

velocity ．

Points S2， M2， B2 are more or less equal with points S4， M4， B4 and both increase towards S3，M3，B3． Therefore， taking S3， M3，B3 as the center vertical plane of the waterway which is referred to as the experimental center line， it can be observed ，that the i／一homogeneous speed fields are symmetric． This

(7)

E． FLoREs and K． NAKAsAI ： Flow Dlstribution of Water Tank 109

disregards the area between S2， iM2， B2 and Si， Mi， Bi where there is a drop of flow speed towards the latter．

．Reference pointレーπto IX一一Since the reference points V to IX fall on a straight waterway， there is not much difference in their cross sectional structure・

This is further clarified by the dominance of almost straight isovels shown on the horizontal cross section of the distribution of flow speed （Fig． 8） covering the above reference points．

30 co 〜 40

− 一

一ーー−ーー− 一

Fig． 8．

so ^− 〜 40 ︑〜〜 ao

一 1

B

Horizontal cross section of flow speed distribution from reference points V to IX；（S） surface water，（M） mid−water and （B） bottom water．

There is still a noticeable drop of flow speed in the一 area of the inner wall rendering it not favorable for model net testing． On the other hand， the difference is not so much in the area about the experimental center line， and this area was decided to be appropriate for model net testing． ln this area， at 40 cm／sec mean velocity， the average deviation is 6 cm／sec．

As the model net is to be placed in between the reference points VI and VTJ，

special attention was given to these points．

Reference points X and XI 一一 With the water passing through another set of angular conduit plates， point B3 has lesser velocity than M3 throughout the series．

An example at motor scale 6．5 is shown in Fig． 9． ln the figure， the isovel pattern from S3 to M3 is continued down to B3 with maximum velocity at the surface and minimum at the bottom． As mentioned ealier， water at S level

(8)

Fig． 9． Vertical cross section of flow speed （cm／sec） distribution at reference points X and XI．

encounters less frictional force and that at deeper layers experiences more ．frictional force because of the bottom and the walls．

For the second series of experiments，

there iS a general decrease of flow speed with the use of the meshed wires． At the same time， there is also a decrease in the flow speed difference between points of measurement showing an improvement in the flow speed distribution．

Since the model net is to be placed at

15cm from the bottom supported by a

plastic board， a closer analysis of this area is necessary using the isovels of the hori−

zontal flow distribution．

With the combined B3 and M3 points of VI and VIJ as the experimental center line， the differences （D） in flow speed on both sides towards the vertical planes of B2， M2 and B4， M4 are shown below．

The center of the model net coincides with the experimental center line． The main part of the model net will experience the flow speed about the experimental

Table 1． Velocity difference in cm／sec of B2−B3−B4 and M2−M3−M4，

Motor

scale R ^B2 D ^B3 D ^B4

4

5

6

7

8

9

9．8

IH 111 111 111 111 111 111VV VV VV VV VV VV VV 17 17 22 22．5 27 28 34 32 39 37 44 42 46 48

−⊥−⊥ ﹁D だ0

9臼雪⊥ 21ゐ04 34 14 32

18 18 24 24 29．5 29 34 36 42 41 45 46 49 50

19臼 9︺9臼

9自324 54 24 34 5 22 33 33 44 441可⊥ 9臼9臼 76 22 76 22 77 32 66

M2 D M3 D M4

17 17 23 22．5 27

27．5

34 34 38 37．5 43 45 48 48

9臼9臼 5 ﹇O

﹂⊥−⊥ 9﹂9臼9θ3 54441晶 QJ90 99 44 09 67 22 76 11 22 32 33 44 44．55可⊥1⊥ 32 23 43 55 54 55 65 ρ07 9臼− ρ061←1 22 22 −⊥9臼QU∩δ 76 2噌⊥ 7ρ033 44 44

(9)

E． FLOREs and K． NAKASAI Flow Distribution of Water Tank ¹¹¹

center line， and only the anterior portions of both wings are subjected to the flow speeds of M2， B2 and M4， B4・

This setting proved to be satisfactory as shown in Fig． 10 by the generally symmetrical shape of the headline of the model trawl net during the experiments that followed．

欝

・聯

繋

騒，

Fig． 10． Top view of model net under test showing symmetrica l headline configuration．

The authors greatly appreciate the Expenses for executing this study were Education of Japan．

cooperation of Mr． K． Nishimura．

partly defrayed by the Ministry of

Reference

1） KAwAKAMi， T．： Development of mechanical studies of fishing gear． Modern Rishing Gear of the Morld ， Fishing News （Books） Ltd．， London， 175−184 （1968）

2） NARAsAKo， Y． and M． KANAMoRI： A large−sized experimental tank of twin symmetric elliptical circuits． Modern Pishing Oθαrげtゐe Morl〔！， Fishing News（Books）Ltd．，

London， 205−208 （1968）

3） TAuTI， M．： A relation between experiments on model and on full scale of fishing net． Bull． Jap． Soc． Sci． Rish．， 3，（4）， 171−177 （1934）