You are on page 1of 6

Flow Measurement and Instrumentation 65 (2019) 16–21

Contents lists available at ScienceDirect

Flow Measurement and Instrumentation


journal homepage: www.elsevier.com/locate/flowmeasinst

Re-definition of the discharge coefficient of throat-tapped flow nozzle and T


investigations on the influence of geometric parameters

Noriyuki Furuichi , Yoshiya Terao
National Institute of Advanced Industrial Science and Technology, National Metrology Institute of Japan (AIST, NMIJ), Tsukuba Central 3, 1-1-1 Umezono, Tsukuba 305-
8563, Japan

A B S T R A C T

The objective of this paper is to obtain a useful equation which gives a discharge coefficient of a throat-tapped flow nozzle. Experiments for throat-tapped flow
nozzles with several different geometric parameters were performed using highly accurate flow facilities at the National Metrology Institute of Japan (NMIJ). The
effects of some geometric parameters on the discharge coefficient were experimentally investigated. The parameters include the ratio of the diameter of the pressure
taps located at the throat to the throat diameter, the diameter of the pressure taps located at upstream pipe and the roughness of the nozzle surface. Also, the
effectiveness of a flow conditioner prescribed in ASME PTC 6 was examined. Although a limitation to the proposed equations was found, the discharge coefficient is
affected by the geometric parameters smaller than the measurement uncertainty. Finally, new equations for the discharge coefficient, which cover all experimental
results within 0.5%, have been proposed as a set of functions of the Reynolds number.

1. Introduction authors [7–10]. The typical results of the discharge coefficient Cx ex-
perimentally obtained by the authors are shown in Fig. 1, in comparison
Throat-tapped flow nozzles [1–3] manufactured in accordance with with Eq. (1). As the Reynolds number Red increases, the discharge
ASME standard [4], hereafter ASME, are widely used in power plants coefficient tends to increase for Red < 8.0 × 105, then to decrease
not only to evaluate the efficiency of steam turbines but also to measure slightly for 8.0 × 105 < Red < 2.0 × 106, and to increase again for
the feedwater flow rate in nuclear power plants, which means this type Red > 2.0 × 106. The general trend is consistent between Eq. (1) and
of flow nozzle is one of the most important sensors for controlling our experimental data, however, the increasing rate of the measured
power plants. ASME defines the following standard equation, which discharge coefficients is clearly higher than that of Eq. (1) especially at
was theoretically derived by Murdock et al. [5], to express the dis- the higher Reynolds number (Red > 3.0 × 106). Thus, if kT in Eq. (1) is
charge coefficient CPTC6 for a throat-tapped flow nozzle, estimated from results of actual calibration performed at lower Rey-
0.8 nolds number, discharge coefficient obtained by Eq. (1) can be sig-
0.185 ⎛ 361239 ⎞
CPTC6 = kT − ⎜ 1− ⎟ nificantly smaller than the actual calibration results at higher Reynolds
Red 0.2 ⎝ Red ⎠ (1) number. From the experimental results shown in Fig. 1, the maximum
where Red is the Reynolds number based on the diameter and the ve- value of the difference is up to 0.5%.
locity at the throat of the nozzle. The first term of the right-hand side is As reported in the series of previous papers [7–10], we have studied
a correction coefficient. The nominal value of kT is 1.0054 but can be the physical meaning of the characteristics of the discharge coefficient
substituted by a value obtained from results of an actual calibration. as a function of the Reynolds number. Especially, we focused on the
The second term of the right-hand side is obtained by theoretical ana- pressure measurement error of the wall tap to clarify the source of the
lysis of the boundary layer. ASME states this equation is applicable over deviation of Eq. (1) from the measured data. Finally, we found that the
a certain Reynolds number range, which is inconsistency is caused by the static pressure measurement error at the
1.0 × 106 < Red < 5.0 × 107. Even if a value of kT is determined from pressure tap located in the throat area. In Eq. (1), the static pressure
calibration data performed at a lower part of the mentioned Reynolds measurement error is assumed to be independent from both the Rey-
number range, the discharge coefficient can be estimated using Eq. (1) nolds number and the diameter of the pressure tap, based on the results
over the whole Reynolds number. by Shaw at 1960 [11]. However, recent experiments, which were re-
Recently, significant inconsistency between some actual calibration ported by McKeon and Smits [12] and also the authors [9], indicate that
results and Eq. (1) was reported by Reader-Harris et al. [6] and the the static pressure measurement error varies with Reynolds number and


Corresponding author.
E-mail address: furuichi.noriyuki@aist.go.jp (N. Furuichi).

https://doi.org/10.1016/j.flowmeasinst.2018.11.007
Received 30 May 2018; Received in revised form 6 November 2018; Accepted 10 November 2018
Available online 14 November 2018
0955-5986/ © 2018 Elsevier Ltd. All rights reserved.
N. Furuichi, Y. Terao Flow Measurement and Instrumentation 65 (2019) 16–21

Table 1
Geometric parameters of 6 flow nozzles examined in the present and previous
experiments.
Nominal diameter D (mm) 100, 200, 350
Throat diameter d (mm) 50, 99, 167
Diameter ratio β app. 0.5
Throat-tap diameter dT (mm) 2, 3.5, 4, 5, 6, 7
dT/d 0.012 – 0.1
Upstream-tap diameter dU (mm) 2, 4, 5
Surface roughness Ra (μm) 0.10, 0.80
Rt (μm) 0.60, 2.5

parameter and the examination methods are explained in the following


sections from 2.2 to 2.5.

Fig. 1. Experimental results for the variation in the discharge coefficient with 2.2. dT/d effect
Red [8]. d is the diameter of throat and dT is diameter of pressure tap at throat.
As mentioned in the introduction, we proposed a set of ideal
equations of the discharge coefficient of the throat-tapped flow nozzle
with the tap diameter. Based on those experiments and investigations,
in the previous paper [8]. Each equation covers different Reynolds
we have proposed ideal equations, which express the discharge coeffi-
number range. The equation for highest Reynolds number region is
cient of the throat-tapped flow nozzle, in the previous paper [8].
given by the following.
However, the manufacturing process of the pressure tap is critical,
0.8
because a very small burr or roundness of the edge of the wall tap can 0.255 ⎛ 400000 ⎞ dT
CF = 1.0042 − 1− + {0.0746ln(Red ) − 0.9051}
influence the discharge coefficient. In addition, a throat-tapped flow
⎜ ⎟

Red 0.2 ⎝ Red ⎠ d (2)


nozzle has many geometrical parameters such as the diameter of wall
6
tap, the diameter ratio, the surface roughness and so on. Since the This equation is used for Red > 3.0 × 10 . The discharge coefficient
proposed equation in the previous paper is derived from the results is the function of Reynolds number and dT/d. In the previous paper, the
using only two different nozzles, the applicable range for those para- examined range of dT/d was up to 0.035. To obtain the limitation of dT/
meters was not defined. In this paper, the influence of the such para- d range, the examinations for larger dT/d are performed in this paper.
meters is investigated in detail and the applicable range of the equation
for each parameter is clarified. Based on these experimental results, the 2.3. Diameter of upstream tap
valid equation of the discharge coefficient of the throat-tapped flow
nozzle for actual use is proposed. The discharge coefficient for the throat-tapped flow nozzle is af-
fected also by the diameter of the pressure taps. This is caused by the
2. Experimental parameters static pressure measurement error as reported in the previous paper [6].
In general, the static pressure measurement error increases with the
2.1. Examined flow nozzles diameter of the pressure tap as given in the following [11],

e d
The schematic figure of the throat-tapped flow nozzle is shown in = f ⎛Ret, T ⎞
Fig. 2 and the geometric parameters examined in a series of experi- τw ⎝ d⎠ (3)
ments and present experiment are shown in Table 1. We have examined where, e is the static pressure measurement error, τw is the wall shear
three nozzles in the previous experiments and newly examined three stress and Ret = dTuτ/ν. Eq. (3) indicates that the static pressure mea-
more nozzles in this paper. These nozzles have different geometric surement error is the function of the wall shear stress on the pressure
parameters. The parameters are the diameter of the nozzle, the dia- tap. Due to the difference of the wall shear stress between the throat-tap
meter of the pressure taps, the roughness of the nozzle. A pressure tap and upstream-tap, the differential pressure is influenced by this effect,
located at the throat is called a throat-tap and the tap at upstream is called as tap effect hereafter. The differential pressure decreases with
called an upstream-tap, hereafter. The nominal diameter D of the nozzle the diameter of the throat-tap, resulting in an increase in the discharge
is from 100 mm to 350 mm. The throat diameter d is selected so that the coefficient. Since this phenomenon applies not only to the throat but
diameter ratio β should be approximately 0.5. The details of each also to the fully developed flow [11,12], the influence of the upstream-
tap diameter to the discharge coefficient of the throat tapped flow
nozzle is examined in this paper.

2.4. Surface roughness of the nozzle

The surface roughness is one of the important parameters for the


discharge coefficient of the nozzles. Wang et al. [13] reported that the
discharge coefficient of the sonic nozzle decreases with increasing of
the surface roughness of the nozzle. The similar influence might be
observed for the throat-tapped flow nozzle. The development of the
boundary layer depends on the surface roughness. In general, the larger
the surface roughness is, the faster is the development speed of the
boundary layer. To avoid the influence of the surface roughness, the
wall surface should be hydraulically smooth as defined in PTC 6. To be
hydraulically smooth surface, the roughness factor ks+, which is de-
Fig. 2. Schematic of flow nozzle. fined as the following, should be less than 5 [14],

17
N. Furuichi, Y. Terao Flow Measurement and Instrumentation 65 (2019) 16–21

k s+ =
u τ ks pressure. The differential pressure Δp was measured using digital
υ (4) manometers. The start and stop times for the differential pressure
where, ks is the height of surface roughness, uτ is the friction velocity measurement were precisely synchronized with the duration of the
and ν is the kinematic viscosity. Using the relation between the friction reference flow rate measurement. The sampling rate for the digitized
velocity uτ and the friction factor λ, differential pressure was between 0.5 and 1 s. The volumetric flow rate
was obtained using reference flowmeters, a weighing tank system, and
uτ = λ /8 Ub (5) a prover system. The density was corrected using a density meter.
Further measurement details can be found in previous reports [7].
where, Ub is the bulk velocity. The height of the surface roughness
The standard uncertainty (k = 1) of the volumetric flowrate are
should be satisfied the following relation to be a hydraulically smooth
estimated to be from 0.02% up to 0.05% depending which facility is
surface,
used. The uncertainty of the geometric parameters, d and D are defined
ks 10 2 by actual length measurements. The density of the water is checked
<
d Red λ (6) using the density meter and the uncertainty is from this measurement.
The uncertainty of the differential pressure measurement is estimated
Here, the following relation reported recently by Furuichi et al. [15] from the calibration result of the digital manometer. Finally, the un-
is used to estimate the friction factor, certainty of the discharge coefficient is estimated to be 0.08–0.12%
1 depending on the measurement condition.
= 2.092log(Red λ ) −1.176
λ (7) In the experiments, the ranges of the flow rate, the water tem-
perature and Reynolds number differed among the examinations but, in
and the relation between ks/d and the Reynolds number is obtained total, the flow rate and temperature were changed from 0.006 to
with the following equation, 0.61 m3/s and from 13 to 70 °C, respectively. The Reynolds number
ks range based on the diameter and the velocity at the throat was from
< 43.34Red−0.92
d (8) 1.5 × 105 to 1.1 × 107. The temperature fluctuation during one mea-
surement campaign was less than 0.2 °C, and the pressure fluctuation
In this paper, this limitation is validated experimentally by using the
was less than 0.01 MPa.
same size of nozzles but with different surface roughness as given in
Table 1.
4. Experimental results

2.5. Flow conditioner 4.1. Limitation of dT/d

ASME defines that the flow conditioner must be installed upstream The experimentally obtained discharge coefficients, Cx, and the
of the flow nozzle. To avoid the influence of the upstream pipe layout, calculated values of equations given by the previous paper [8], CF, are
the flow conditioner is generally useful. However, as shown in ISO- shown in Fig. 4. For visibility of the figure, error bars are shown only
5167 [16], the requirement of the upstream straight pipe length from a for dT = 2 mm data. The dashed line is from PTC 6. The solid lines were
single elbow to the nozzle inlet is 14D for β = 0.5. Since the straight obtained by substituting the diameter of the pressure tap into the
pipe length of the throat-tapped flow nozzle is app.20D, sufficiently equations reported in the previous paper [8]. The diameter of the throat
long pipe is installed upstream of the nozzle. In terms of the pressure d is 100 mm in this experiment.
loss, the necessity of the flow conditioner should be checked. In this The discharge coefficient obtained in this experiment increases with
paper, the examination was carried out with and without the flow the diameter of the pressure tap, and the results reveal the tap effect.
conditioner. Moreover, a half-moon plate, which blocks the half area of The behavior of the discharge coefficient, where it increases with the
the pipe, as flow disturber is installed to generate a velocity field similar diameter of the pressure tap, agrees well with the results of the previous
to that downstream of a single elbow. The installation position is the study [8]. The experimental results show good agreement with CF for
same as for the flow conditioner as shown in Fig. 2. Red < 5.0 × 105 and dT = 2 mm. However, for dT = 3 and 4 mm, CF
and the experimental results are not in agreement. Furthermore, for
3. Experimental method Red > 5.0 × 105, the experimentally obtained discharge coefficients do
not agree with CF for any tap. This indicates that equation given CF have
3.1. Experimental facilities a limiting dT/d value. In the previous papers [7–10], the examined dT/d
values were less than 0.036. From the present and previous results, the
The flow sheets for the two experimental facilities used in this study limiting value of dT/d is expected to be approximately 0.04.
are shown in Fig. 3. The experiments in the low flow rate region were
performed at the water flow facility, and those at high flow rates and 4.2. Influence of upstream-tap diameter
temperatures were performed at the High Reynolds Number Actual
Flow Facility (Hi-Reff). These facilities serve as the national standard of The experimental results for dU = 2 and 5 mm are compared in
water flow rate in Japan, and the flow rate extend uncertainties are Fig. 5. For visibility of the figure, error bars are shown only for dT
very low: 0.040% (k = 2) for the water flow facility and 0.10% for Hi- = 2 mm, dU = 5 mm data. As the diameter of the pressure tap at the
Reff. The details of these facilities can be found in previous papers upstream pipe decreases, the differential pressure is expected to in-
[17,18]. crease. As shown in the figure, the discharge coefficient increases with
decreasing diameter, and the difference between the results from the
3.2. Experimental methods and uncertainty different dU is very small (less than 0.02%). As reported in the previous
paper [9], the pressure measurement error is related to the wall shear
The experimental discharge coefficient Cx of the throat-tapped flow stress. The wall shear stress on the pressure tap at the upstream pipe is
nozzle is given by smaller than that on the throat. The results shown in Fig. 5 support this
q ρ conclusion.
Cx = 8(1−β 4 ) The uncertainty in the measurement of the discharge coefficient was
πd 2 Δp (9)
estimated to be approximately 0.08% with a coverage factor of 2 in this
where, q is volumetric flowrate, ρ is density of water, Δp is differential experiment. Combining the uncertainty of 0.02% from the tap diameter

18
N. Furuichi, Y. Terao Flow Measurement and Instrumentation 65 (2019) 16–21

Fig. 3. Experimental facilities. (a) Water flow facility. (b) Hi-Reff.

of the upstream pipe with the measurement uncertainty yields an es-


timated combined uncertainty of 0.08%. Regarding the uncertainty in
the measurement, it was concluded that the influence of the tap effect at
the upstream pipe on the discharge coefficient is negligibly small.

4.3. Influence of roughness

Two nozzles with different roughness were examined to evaluate


the maximum permissible value of the surface roughness. The first is
the nozzle, R1, which has a roughness of Ra = 0.2 µm, Rt = 0.8 µm and
was investigated in previous studies by the authors [7,8]; the other, R2,
is the nozzle, which has a roughness of Ra = 0.8 µm, Rt = 2.8 µm. The
experimental results for these nozzles are shown in Fig. 6 for different
throat-tap diameters. For visibility of the figure, error bars are shown
only for R1, dT = 2 mm data. In the transitional region
Fig. 4. Experimental result for large dT/d. The diameter of the throat d is
(1.0 × 106 < Red < 3.0 × 106), the influence of the roughness is not
100 mm.
investigated. However, especially for the taps with large diameters, the

Fig. 5. Influence of the diameter of upstream- tap. The diameter of throat d is Fig. 6. Influence of the surface roughness of the nozzle. The roughness of R1 is
100 mm. Ra = 0.2 µm (Rt = 0.8 µm) and R2 is Ra = 0.8 µm (Rt = 2.8 µm).

19
N. Furuichi, Y. Terao Flow Measurement and Instrumentation 65 (2019) 16–21

discharge coefficients of the two nozzles are observed to deviate from


each other at higher Reynolds numbers, Red > 8 × 106. The discharge
coefficient is found to decrease with increasing roughness. This result is
consistent with the physics of the boundary layer. The thickness of the
boundary layer increases with increasing roughness. This phenomenon
results in a reduction of the discharge coefficient. However, the dif-
ference is not large compared with the measurement uncertainty. The
difference between the discharge coefficients for the two nozzles is
0.02% on average and reaches a maximum of approximately 0.05%.
To obtain the Reynolds number limitation, under which there is no
effect from the surface roughness, the roughness height ks in Eq. (8) is
estimated. Since the throat diameter d is 167 mm in this experiment, it
is estimated that the value of Red to be 3.6 × 107 when Ra is used, Red
= 9.3 × 106 when Rt is used for the rough nozzle. Based on ISO defi-
nition for the surface roughness [19], Rt is the summation between the
maximum height and minimum height and it is reasonable to consider Fig. 8. Discharge coefficients measured in all experiments.
as ks. This consideration is supported by the experimental result that the
influence of roughness can be observed Red > 8 × 106.
supported by the physical considerations of the flow field. However,
there are some cases, where the equations for CF cannot be adopted as
4.4. Flow conditioner shown in the previous section, although the deviation is relatively
small. Especially the equations have a limitation for the value of dT/d,
The discharge coefficients for each upstream condition, with flow which must be less than 0.04.
conditioner (FC), without flow conditioner and with the half-moon All data obtained in this study is plotted in Fig. 8 along with the
plate (HP) are shown in Fig. 7. For visibility of the figure, error bars are standard equation from ASME. The Reynolds number range is from
shown only for data with flow conditioner. The results for dT = 2 mm 5.0 × 104 to 1.5 × 107. The figure clearly indicates again that the
and 6 mm are shown. The discharge coefficient with and without flow standard curve defined in ASME does not reflect the actual discharge
conditioner agree well each other. On the other hand, the discharge coefficient for the throat-tapped flow nozzle. The deviation in the high
coefficient in the case of the half-moon plate is slightly deviated from Reynolds number region, which is the most common operational re-
the result for without it for both taps. The discharge coefficient is gion, was given particular focus in this investigation.
app.0.06% smaller than without case. This deviation is also not large The authors propose a new recommended curve to determine the
compared with the measurement uncertainty. When the upstream discharge coefficient for the throat-tapped flow nozzle. The equations
length is over than 22D, the influence of the upstream condition ap- for the proposed curve, plotted in red in Fig. 8, were developed by the
pears to be not necessary to consider. This is reasonably equivalent with authors in the previous study [8]. The parameter dT/d was selected to
the requirement of straight pipe in ISO-5167. be 0.024 to minimize the deviation from the equation, and they are
given by the following.
5. Discussion 8.41
CF = 1.0042− for Red <1.3 × 105
Red 0.5 (10)
The authors have conducted a large number of experiments on
throat-tapped flow nozzles as shown in the previous [7–10] and the CF = 0.9558−
8.41
+0.00492ln(Red ) for 1.3 × 105 < Red < 4.0 × 105
present paper. The nominal diameter investigated in these papers Red 0.5
ranged from 100 to 350 mm, and the throat diameter was from 50 to (11)
165 mm. The diameter of the throat-tap and upstream-tap ranged from
8.41
2 to 7 mm and from 2 to 5 mm, respectively. The range for dT/d was CF = 1.0090− for 4.0 × 105 < Red < 8.0 × 105
from 0.012 to 0.1. The maximum roughness was Ra/d = 4.8 × 10−6 Red 0.5 (12)
(Rt/d = 1.6 × 10−5). The authors have reported the ideal equation, CF, 0.8
0.255 ⎛ 400000 ⎞
a set of five equations, which represents the experimental discharge CF = 1.0090− ⎜ 1− ⎟ for 8.0 × 105 < Red < 3.0 × 106
Red 0.2 ⎝ Red ⎠
coefficient over wide Reynolds number range [8]. Those equations are
(13)
0.8
0.255 ⎛ 400000 ⎞
CF = 0.9823− ⎜ 1− ⎟ +0.0018ln(Red ) for 3.0 × 106 < Red
Red 0.2 ⎝ Red ⎠
(14)
The proposed equations cover all presently considered experimental
data for the discharge coefficient within an error of ± 0.5%.

6. Conclusion

Throat-tapped flow nozzles were examined to determine the influ-


ence of several parameters on the discharge coefficient. First, the size
effect was investigated using a small flow nozzle with a nominal dia-
meter of 100 mm. The discharge coefficient was in good agreement
with the ideal equations, CF [8] at low Reynolds numbers with dT/
d < 0.04. For high Reynolds numbers and large dT/d, the experimen-
tally obtained discharge coefficient did not agree with either CF or the
Fig. 7. Influence of flow conditioner and flow disturbance. standard equation given in PTC 6 [4].

20
N. Furuichi, Y. Terao Flow Measurement and Instrumentation 65 (2019) 16–21

The effect of the diameters of the throat-tap and upstream-tap was nozzles, J. Fluid Eng. 100 (3) (1978) 265–274.
investigated. The considered diameters were 2 and 5 mm, and the dif- [4] ASME PTC6, Steam Turbines, Performance Test Codes, 2004.
[5] J.W. Murdock, D.R. Keyser, Theoretical basis for extrapolation of calibration data of
ference between the discharge coefficients for these two diameters was PTC 6 throat tap nozzles, Trans. ASME J. Eng. Gas. Turbines Power 113 (1991)
less than 0.02%. Thus, the influence of this factor is negligibly small. 228–232.
The influence of the roughness of the nozzle surface was also in- [6] M. Reader-Harris, J. Gibson, D. Hodges, I.G. Nicholson, R. Rushworth, The per-
formance of flow nozzles at high Reynolds number, in: Proceedings of FLOMEKO
vestigated. Throat-tapped flow nozzles with Ra = 0.2 and 0.8 µm 14, Johannesburg, South Africa, 2007.
(Rt = 0.8 and 2.8 µm) were examined. The maximum difference be- [7] N. Furuichi, K.H. Cheong, Y. Terao, S. Nakao, K. Fujita, K. Shibuya, New discharge
tween the discharge coefficients for these two roughness values was coefficient of throat tap nozzle based on ASME performance test code 6 for
Reynolds number from 2.4 × 105 to 1.4 × 107, J. Fluid Eng. 136 (1) (2013)
approximately 0.05%. Finally, the influence of the flow conditioner is 011105, https://doi.org/10.1115/1.4025513.
examined. When the length of upstream straight pipe is over 22D, we [8] N. Furuichi, Y. Terao, S. Nakao, K. Fujita, K. Shibuya, Further investigation of
found that the upstream condition has negligible influence on the dis- discharge coefficient for PTC 6 flow nozzle in high Reynolds number, J. Eng. Gas.
Turbines Power 138 (2016) 041605–1-11.
charge coefficient.
[9] N. Furuichi, Y. Terao, Static pressure measurement error at a wall tap of a flow
As a summary of the presently reported experiments on the throat- nozzle for a wide range of Reynolds number, Flow Meas. Instrum. 46 (2015)
tapped flow nozzle, new equation for the discharge coefficient are 103–111.
proposed as useful model. The equations can represent the discharge [10] N. Furuichi, L. Cordova, T. Lederer, Y. Terao, Comparison of high temperature and
high Reynolds number water flows between PTB and NMIJ, Flow Meas. Instrum. 52
coefficient by all experiments within 0.5% as a function of the Reynolds (2016) 157–162.
number. [11] R. Shaw, The influence of hole dimensions on static pressure measurement, J. Fluid
Mech. 7 (1960) 550–564.
[12] B.J. McKeon, A.J. Smits, Static pressure correction in high Reynolds number fully
Acknowledgements developed turbulent pipe flow, Meas. Sci. Technol. 13 (2002) 1608–1614.
[13] C. Wang, H. Ding, Y. Zhao, Influence of wall roughness on discharge coefficient of
Part of the experimental data for the discharge coefficient for the sonic nozzles, Flow Meas. Instrum. 35 (2014) 55–62.
[14] H. Schlichting, K. Gersten, Boundary Layer Theory, 8th Revised and Enlarged
flow nozzle was obtained in collaborative work with Flow Engineering Edition, Springer, 2000.
Co., Ltd. The authors would like to thank Mr. Kazuo Shibuya and Mr. [15] N. Furuichi, Y. Terao, Y. Wada, Y. Tsuji, Friction factor and mean velocity profile
Keiji Fujita of Flow Engineering Co., Ltd., for assisting with the ex- for pipe flow at high Reynolds numbers, Phys. Fluids 27 (2015) 095108, https://
doi.org/10.1063/1.4930987.
periment and Dr. Shinichi Nakao of FlowCol for fruitful discussions. [16] ISO 5167, Measurement of fluid flow by means of pressure differential devises in-
serted in circular-cross section conduits running full, 2003.
References [17] N. Furuichi, H. Sato, Y. Terao, M. Takamoto, A new calibration facility of flowrate
for high Reynolds number, Flow. Meas. Instrum. 20–1 (2009) 38–47.
[18] N. Furuichi, Y. Terao, Y. Tsuji, High Reynolds number experimental facilities for
[1] G.W. Hall, Application of boundary layer theory to explain some nozzle and Venturi turbulent pipe flow at NMIJ, Progress in Turbulence VII, Springer, 2017, pp. 89–94.
flow peculiarities, Trans. IME Lond. 173 (36) (1959) 837–870. [19] ISO 4287, Geometrical product specification (GPS). Surface texture. Profile method.
[2] K.C. Cotton, J.C. Wescott, Throat tap nozzles used for accurate flow measurements, Terms, definitions and surface texture parameters, 1997.
Trans. ASME J. Eng. Power (1960) 247–263 (October 1960).
[3] R.P. Benedict, J.S. Wyler, Analytical and experimental studies of ASME flow

21

You might also like