292. Discussion of paper by Yarnell and Nagler on effect of turbulence on cup and propeller meters, 1930

(1)

(2)

IN YOUR REPLY PLEASE REFER TO FILE No.

-UNITED STATES DEPARTMENT OF AGRICULTURE BUREAU OF PUBLIC ROADS

DIVISION OF AGRICULTURAL ENGINEERING

H

r. Carl Rowwer,

A soc. Irr1g.

~ngr.,

Burea.u of Public Roads,

Ft. Collins, Colo.

Dear Mr.

Rohwer.

Iowa City, Iowa.

January

22, 1930

.

I

have

you~r

letter of January

17 together w1th copy of your discussion of our paper on

the current meter tests

.

Professor Nagler and

I

are

-very glad to have the opportunity to read your discussion

before apoearing in the Proce dings

.

Your

points are well

ta~en

and we should correct

some of the headings of the tables

.

7e

are glad tbat

you are including some tables of your own work in your

discussion

.

We want to thank you for discussing ;ur paper

and to assure

you

that we will be

glad

to have y&ur

discussion appear in the next issue of the Procedd1nga

.

I

am

eeturning the copy of your discussion

.

,.

D. L.

Yarn

Senior Drainage

(3)

( t

\.

• _•

a.

(4)

t

---7, l 1 l~ Q.r " , 1 H.

o

..

( \' J·l$t'j o.o2~1

v

2.116 _"

~},.

o.o

7 3

.o

f"\

•

1 b · l

u

.o

,... #

31,

as

J~o

1&.

l9Z

v.

_2·1

.a

+

o.ozo

lin

Shutt

1 ·t

uoo

.y

f 9 --3' _~

*

o.o

2·4

. l .~ 4

.o

\')

vat

...

\t 1 .. t l l

• lw

A;rU !L .. l

.

n

.o

1

_•

1;1 ....

v

.2. 72 P. o.o~ _~ ( lt

..

0 _r ~

•

) f

(5)

at th" l."ati

t

1 b t J .. ~ t di

t:

(il'f>ll

,

d - ln. 2

-non

1t

on

:rt

OolliM 1 Colo. loe1tt;'/ th tO low-

v

,0(!

ty

I

t

(6)

c 1

iW!

t t b u. I 1 8 o l"t in "

11

n _e

•

t t r

.

tio

or

n on

t

e OCU.a.' C'!/

que

0 t 1 t

co

1

(7)

c

ur

·

-1. ~

..

.

?

'!

• u

.

o

"t h. s

taJ.

-

e.

t rli~· ~ 1 t

e

c~·

th

a1

eo

of

1':'$ 1 t I ..

-

...

(8)

c

;rl

_'~'• 00

.

,

_·

_1·

_•

6

-

_•

'·

1~1

1.

1 1·"

-::

i!

_•

7

.,

e

.

+

..

o

,

•

1 '

1

,,'1

-

.

"

... •

,

.a

(

-

.~~

-

_•

fl . t;.l

'

'·

.'1'1

•

4

1.

,/ 7.982 2.091 1. ?51 29.22

...

• -o,

• •l

fi.O

a.

"'

(9)

eek:.e b il~ ·to

t"

e p n s o 1 e I')

_g1

l l : ang .;

or

1'' ~.s 1th the .. 1 1V 1:1

t

h ®vint mt met r "'h ... n

or eott on.

tJ · !.tl t the

ith t l

n

th-· ., di !d d. •1

ox

th ll b:rs ow 'the l l

e

ion .• r

:cy-.

th th :mp :t · s ti:>t

c

· ory

d

d1

t 0

U1

ion

to

•

(10)

bffect of 1_{furbulence on the Registration of Current !<J!eters.}

Discussion

Carl Rohwer, Assoc. L. A.-·n. Soc. C. E. (By letter)*; The current meter is

frequently theonly available means of measuring the discharge of natural streans

or artificial c~annels,and any information that bears on this subject is always

we~comed by those who are concerned with the measurement of water. The authors

of this paper have carried on e~ensive experiments that show the action of

cur-rent meters under extren:e conditions, as well as the condi tiona ordinarily met

with in practice, and are to be commended for their work. Their summary of

the comment of other investigators on the behavior of various types of current meters under different conditions is also valuable.

1ne authors men

tio~that

the meters used in the

ex~riments

were not

rated at the time of the tests because only the relative values of the

observa-tions whe!f compared with themselves were to be considered in the experiments.

t:J

In Table

II,

however, the velocities indicated by the meters are compared with

the veolcity computed from the weir discharge .. Obviously in this case any error

in the meter rating will be included in the deviation due to the obstruction in

the channel • The error due to the inaccu1~cy in the rating of the meter is

probably small in corqparison with the differences observed; nevertheless, there

is a definite change in the rating of meters as is shown by

Table

1f~

· hich

gives

A

-

-the ratings of a snn 11 Price electric meter extending over the period from 1919

to 1929. These ratings were all made under the same conditions at the rating

station of the hydraulic laboratory at Fort Collins, Colo. It wi 11 be noted

(11)

2-'

there is a definite change between ratings. This is not due to inaccuracies inherent in rating because check ratings in 1921 and 1929 are ~lmost identical.

~

Almost without exception the variation in mean velocity ~d by

dif-~

ferent obstructions in Table I I is greater at the low velocity than at the high

velocity. This seems unusual. Ordinarily, the faster the water flows the greater is the disturbance caused bY. the obstruction. In the data shown in Table II, this seems to hold true only in the case of the submerged weir.

The

data~

in Figures

12~d

1ishow that under some of the extreme conditions of the test, single-count meters give very erroneous results. This does not mean that errors of this magnitude are to be e~ ected under conditions met with in practice. It is doubtful if the velocity would ever rewerse at a point in any section that would be chosen for a current meter measurement

It is true that the water may flow backwards near the banks at certain stages at some rating statio~ but.as long as the water rerains at the same stage, the water will continue to flow in the s·ame direction. The records of all the meters show that a variation of 50;0 in the velocity has practically no effect on the accuracy of registration.

in

Table The heading indicates that the values given .1

are the "Ratios of velocity recorded by meter at various angles shown as a percentace of cosine values of velocity recorded by meterwith axis parallel to current. n These values are apparently the errors in the velocities and

not the ratios of the velocities. The same criticism holds for Figures 14 and

~5 also.

As shown in Table III, none of the meters, except the small Ott, gives very sa tis factory results when measuring the axial component of the velocity !bf the

water~ the meter at an angle in the horizontal plane. The Price meters

""c:.o..~ ll r

e-all over-register because they ifi:aaase the velocity in the direction that the

(12)

,

\ 0

~-\ \

(13)

3-~"-+e-r

~ is rroving and not along the axis of the meter. The errors in tho Price

meters shown in the table, are the deviations due to multiplying the axial

ve-loci ty by the cosines of the angle at which the water appro aches the meter. This

~~

~3)

r~"'~c.c.-is proved by the following tablefWh1ch show~ or each angle and each ~e meter

the reduction of the velocity c:Used by multiplyingjhe cosine of the angle

~nd

~g.,,, ~~~~~ ~b~"·'~

.. ,

.

the ~registration of the meter~ In the com}Xlrison the means of' the right

\f;..a.,

and left deviations are used.

1\

Horizontal Angle Degrees

.e~ction

in

ve~city i

~-

~1_02_0-2_5-3_0-35-4_0-4~5

pe cent~ due to

multiply-i Khe cosine of the angle

l.

a

6 .G

9.4·

O~erregistration

of meter

~

j'

i per cent

UJ'-

'~

tfot'q,cu e., d w \'\·11 /

Q"'o\Cl\ o.•1'W\~o:t.W'\-\o-

-1

te,\oc,~'~\~-~- • -

-Price #211,880 -·- - 1.4 4.3 · 8.? ·

13.4 18.1 23.4

.

; 12.8 1§.5 26.6 36.9

Price accoustic

~

4.7 •

e.o

·

12.4" 18.2. 25.2 34.8"

U.S.G.S. lmprcwed Price 11.7' 18.0 25.4 36.2

the meter apparently

~ '? "-\-- "· ... '0

has v ,ry accuracy of ,until it exceeds

ilV

degrees. This characteristic of the Price meters has its advantages as well

as its disadvantages, because for this reason it is not necessary to hold the

meter exactly parallel to the axis of the stream in order to rr.ake the meter

register accurately. '11

1;lis is not true of the propeller meters.

fuen the vrater approaches the meters at a smll horizontal angle, all

the meters register the axial component of the velocity with a fair degree of

accuracy, as is'shown by Table III. This is also shown by the accompanying

(~"e)

"

tableAwhich g1ves the results of some experiments on cup and propeller meters

tlle.6-~o r e,M e,(\

+-using different methods of n:ove!ilfaPt, made at the Bellvue hydrauli~laboratory

(14)

4-in an 8-foot rectangular channel and checked by a standard 10-foot Francis weir

w

~\"'0

constructed according to the plans of the wier used in the L::>well ITyarraulic

Ex~riments. Each meter was held so that the horizontal axis of the meter oode

an angle of 10 degrees with the axis of the channel. The results given in

the table show the deviation in the measured d~scharge by each meter when

com-Q...o t"re.c..·h o If\

pared with new discharge, both with and without the cosine eex~ition. The

·~ e..o ~ u ~ c:, 'IV\ e. V\. -\-~ v...> \ -\-\-')

l9:x.-table shows that the results of the

1,siiD.ll Price dectric with tail and the small

Ott with tail are closer without the cosine correction, the results for the Hoff

meter are bettered by intr.oduci:ng the cosine correction and the results for the

small Price electric without t~il are divided.

All the tests reported by the authors show that ~rr:a 11 changes from

the normal cause relatively snall errors in the registration of the meters.

The large errors occur when the meters are subjected to unusual conditions.

~he tests indicate that if the rating station is chosen properly, both the cup and the propeller meters will indicate the velocity with a satisfactory degree