Pressure Loss in Elbow Pipes of Unplasticized Polyvinyl ...

Industrial Health, 1996, 34, 389-401 389

Pressure Loss in Elbow Pipes of

Unplasticized Polyvinyl Chloride

Takeshi IWASAKI and Jun OJIMA

National Institute of Industrial Health, 21-1, Nagao 6-chore, Tara-ku, Kawasaki 214, Japan

(Received June 10, 1996 and in revised form July 25, 1996)

Abstract: In the ductwork of local exhaust systems, 90° elbow pipes (JIS K 6739) are commonly used to alter the direction of airflow; thus, are important components of

polyvinyl chloride (PVC) ducts. Pressure loss in 90° PVC elbow pipes was investigated by measuring static pressure, and the characteristics of airflow was determined. First, a linear decrease in static pressure was observed at points of the downstream side beyond

a distance of 10 times the diameter (lOd) from the flanged round opening of the smooth VU ducts (JIS K6741). The linear decrease was also observed at points of the downstream

side located at distances of greater than 30d from the elbow pipe. Coefficients of loss in the PVC elbow pipes were found to be constant for the Reynolds numbers ranging

from 3.38 x 104 to 5.96 x 105 for all diameters examined, and a chart of pressure loss was constructed with these coefficients. The coefficients of loss in PVC elbow pipes were

not equivalent to those of metal stamped elbows for any R/d. However, the differences in the coefficients between the metal stamped elbow and the PVC elbow were smaller

with larger R/d values.

Key words: Local exhaust system - PVC elbow pipe - Pressure loss - Coefficient of loss - Ductwork design

INTRODUCTION

The friction chart for the straight pipes based on pressure loss is essential when designing a ductwork of local exhaust systems. Elbow pipes of polyvinyl chloride (PVC) are often used for the system to alter the direction of airflow (Fig. 1), so complete data for pressure loss in PVC pipes with elbow pipes is neces- sary. However, no systematic data has been published on the pressure loss in

ductwork using straight and PVC elbow pipes. Sprenger'' and Iwasaki'' reported

Address for correspondence: Takeshi IWASAKI, National Institute of Industrial Health, 21-1, Nagao 6-chome, Tama-ku, Kawasaki 214, Japan Tel: 044-865-6111, Fax: 044-865-6116

390 T. IWASAKI et al.

the pressure loss for straight PVC pipes, which was equivalent to 2/3 of that for

galvanized steel pipes3-5). This study was undertaken to examine the characteristics of airflow in PVC

elbow pipes by measuring pressure loss, and to construct a systematic chart for

pressure loss in PVC ducts with elbow pipes. A suitable radius of curvature of an elbow pipe is generally about 2 times its diameter, because a radius of cur-

vature smaller than this leads to higher air pressure loss and/or collection of dust in the elbow pipes5'. We used the thick-walled PVC elbow and the straight VU

pipes in the present study, because this combination has been widely used in the ductwork of local exhaust systems at workplaces . The radii of curvature of the elbow pipes (JIS K 6739)6' used in the this study ranged from 0 .55d to 1.30d, because of their commercial availability.

EXPERIMENTALS

Commercially available PVC elbow pipes were examined in this study. Figure 2 shows two types of elbow pipes with different radii of curvature , DL type (R/ d = 0.55 - 0.65) and LL type (R/d = 0.84 - 1.30). Nominal diameters of these

Fig. 1. An example of a

pipes.

unplasticized polyvinyl chloride (PVC) duct with elbow

PRESSURE LOSS IN PVC ELBOW PIPES 391

elbow pipes were 50, 75, 100, 150, 200, 250, and 300 mm. Each elbow pipe was

connected with two PVC straight ducts of the same diameter, which were used

for measurement of friction loss in the straight PVC pipes in the previous study''.

Figures 3 and 4 show a schematic diagram and a complete view of the experi-

mental apparatus for the measurement of pressure loss in the elbow pipe, respec-

tively. The experimental apparatus and setup for the measurement of pressure loss

were essentially the same as those fully described in the previous reports''. An

elbow pipe of the DL or LL type, which was connected with both a straight 12

m long VU pipe (consist of 3 parts in series) and a straight 4 m long VU pipe,

was installed in the upstream position in the experimental apparatus.

Measuring points of pressure loss were positioned along the two straight pipes

at both upstream and downstream sides of the elbow pipe. The inlet end had a

flanged round opening. In a preliminary experiment, a PVC elbow pipe with a

nominal diameter of 150 mm was used to determine appropriate positions for the measurement of static pressure inside the two straight ducts. Two measuring points,

Ps, and Ps, (Fig. 3), were then positioned and four holes of measuring 2 mm in

diameter were made for measurement of static pressure in each straight duct. The

static pressure was measured at these two points with a Betz-type or a digital

Fig. 2. Profile of the DL and LL types of PVC elbow pipes used in this study.


manometer.

The average air velocity in the duct was calculated by velocity pressure read-

ings Pv at two different positions, i.e., Ps, and the 2.5d downstream position from

the precision honeycomb (1© in Fig. 3). The velocity pressure (Pv) was measured

by round duct 10-points Pitot traverse technique with an NPL standard Pitot-tube.

The velocity (V) is given by the following equation.

Fig. 3. Schematic diagram of the experimental apparatus.

Fig. 4. Complete view of the experimental elbow duct and the measurement system.

( V

where g and y are

of air (1.2 kg/m3),

giv The airflow rate the duct.

Q=A.V

Static pressure behind

Figure 5 indicates th

the opening of the LL

different airflow veloci

airflow constant in the

both the upstream and

The first linear decreas

from the round opening

PRESSURE

_

2 g/y) . Pv

the gravitational

respectively.

Q is en by

LOSS IN PVC ELBOW PIPES

acceleration (9.8

multiplying V by

m/s2) and

the cross-s

RESULTS

the duct end of the flanged round ope e change of static pressure depending o

type PVC elbow pipe (nominal diamet ties, and shows the Anlaufstrecke, i.e., it

duct. A linear decrease in static pressu

downstream sides of the elbow for ever

was observed at downstream points mo e

of the VU pipe (d = 154 mm). The

393

(1)

the specific weight

ectional area A of

(2)

Wing

n the distance from

er = 150 mm) at 3

let length to keep

re was observed at

y airflow velocity. re than 10d distant

static pressure de-

Fig, 5. Changes in static pressure with distance from round opening.


crease result agreed with that previously reported'. As shown in Figure 5, the second

linear decrease was observed at the downstream points more than 30d distant from

the elbow for airflow velocities ranging from 10.38 to 32.96 m/s. The static pres-

sure in the duct was found to decrease with an increase in airflow velocity. The

measuring points of Ps, and Ps2 were then located behind the Anlaufstrecke. Ps,

and Ps, were positioned downstream at 15d distant from the opening, and at 37d

to 86d distant from the elbow, respectively. PVC elbow pipes of the DL type were

found to give the same results as those obtained with the LL type pipes.

Pressure loss and coefficient of loss

It is known that the length of disturbed airflow in the second straight pipe behind

the elbow depends upon the angle or curvature of the elbow. Therefore, in

designing actual ductwork, the curving effects on the Anlaufstrecke complicate the

calculation of pressure loss.

The sum of pressure loss in the elbow pipe (AP) consists primarily of curving

loss and friction loss. However, the difference between the total pressure loss and

friction loss is not strictly equal to curving loss, because the distribution of air-

flow velocity in an elbow pipe is different from that in a straight pipe. In this study, pressure loss in an elbow pipe (AP) is calculated by following

equation, assuming that the temperature of air in pipe is constant and pressure loss

in the elbow occurs along the centerline of curving.

H = PT i - PTZ = (Pvi + Ps~) - (Pv2 + Ps2) ...................... (3)

where H: pressure loss observed in the region between two measurement points,

Ps, and Ps2 (mmAq)

PT,: total pressure at Ps, (mmAq)

PT,: total pressure at Ps2 (mmAq)

Pv1: velocity pressure at Ps, (mmAq)

PV2: velocity pressure at Ps2 (mmAq)

Ps,: static pressure at Ps, (mmAq)

PS2: static pressure at Ps2 (mmAq)

Pv, is equal to P,,2 because the cross sectional areas at Ps, and Ps2 in the two pipes

are equal with each other.

Therefore

H = Ps, - ps2 ......................................................................... (4)

AP = [(Ps, - Ps,) - (friction loss of straight pipes located

at both the upstream and downstream sides of

elbow pipe)] .................................................................. (5)


Further, AP can be calculated from Bernoulli's theorem and expressed as velocity head.

OP = ~•y/2g•V2 = c•Pv .......................................................... (6)

where AP: the sum of pressure loss in the elbow pipe (mmAq)

~: coefficient of loss in the elbow pipe (-)

y: specific weight of air in the elbow pipe (Kg/m3) g: gravitational acceleration (m/s2)

V: average velocity of airflow in the elbow pipe (m/sec) Pv: average velocity pressure of airflow in the elbow pipe (mmAq)

Tables 1 and 2 show the coefficients of loss for PVC elbow pipes derived from the equation (6). The coefficients were approximately constant for all the diam-

Table 1. Coefficient of loss for unplasticized polyvinyl chloride (PVC) elbow pipes (DL type).


eters examined, independent of velocities and flow rates of air in the ducts . Averaged coefficients for different diameters are presented as basic data for de-

signing ductwork of local exhaust systems .

Coefficient of loss and Reynolds number

Due to experimental limitations, we could not investigate characteristics of

airflow in the critical zone which appears between the turbulence flow and tran-

sition flow. In the critical zone, pressure loss is too small to be measured with

conventional manometers. In most cases, the ductwork of local exhaust systems

are designed with turbulent airflow ranging in Reynolds number above 3 x 104 and

the duct with inner diameters of more than 50 mm . Figure 6 shows the relation-

Table 2.

(LL type),

Coefficient of loss for unplasticized polyvinyl chloride (PVC) elbow pipes


ship between the coefficient

lence flows and diameters.

Reynolds numbers ranging

of loss and the

The coefficient

from 3.38 x 104

Reynolds number for various turbu-

of loss is approximately constant for

to 5.96 x 105.

Coefficient of loss and radius of curvature

Figure 7 and Tables 1 and 2 show the relationship between coefficient of loss

and radius of curvature. DL type elbow pipes with relatively small diameters

(nominal diameter = 50, 75, 100, and 150 mm) had been predicted to give smaller coefficients of loss than pipes with relatively large diameters (nominal diameter = 200 , 250, and 300 mm), because the radii of curvature of the small pipes are larger than those for large pipes. However, contrary to this prediction, our re-

sults with DL type pipes indicated the opposite trend, while the data of LL type elbow pipes followed it. In designing ductwork, the pressure loss chart given by

Figure 7 should be applied only to elbow pipes with R/d ranging from 0.55d to 1.30d examined.

Fig. 6. Relationship between coefficient of loss and Reynolds number.


DISCUSSION

In this study, characteristics of airflow in two straight PVC-VU pipes connected with a PVC elbow pipe were measured at upstream and downstream points of the elbow pipe (nominal diameter = 150 mm, Fig. 5). Linear decreases in static

pressure were observed in two steps at the points more than lOd distant from the round opening, and at the points more than 30d distant from the joint of the elbow and straight pipes, respectively, for airflow velocities ranging from 5.49 to 32.96 m/s. The first step, i.e., Anlaufstrecke, observed at lOd in this study, agreed well with the value given in the previous study with a straight PVC pipe2'. Tanaka'' reported that the Anlaufstrecke was observed within 6.25d from the opening with a smooth brass duct (d = 32 mm, L = 5.6 m). Iwasaki8' also reported the Anlaufstrecke as 5d-8d from the flanged opening of steel pipes (L = 38d). The Anlaufstreckes observed in this and previous studies2' were greater than the val-

Fig. 7. Pressure loss chart for unplasticized polyvinyl chloride (PVC) elbow pipes.


ues of 6.25d'' and our previous value of 5d-8d. These discrepant results may indicate that using a longer Anlaufstrecke value like lOd from the round opening of the duct can avoid underestimation of its length, allowing more reliable esti-mation of pressure loss when designing the ductwork for a local exhaust system.

Itoy' reported that the decrease of static pressure behind the elbow pipe was

equivalent to that in the straight pipe at the points greater than 50d distant from the elbow. In this study, measuring points were located at the points 54d-86d distant from the elbow for pipes of nominal diameters ranging from 50 mm to 200 mm. For pipes of nominal diameters of 250 mm and 300 mm, we located measuring points at 37d and 44d. This result may be due to experimental limi-tation, i.e., the upper limit of the length of straight pipe was 12 m which was not long enough for larger pipe diameters, because the Anlaufstrecke becomes longer with an increase in the diameter. However, the coefficient of loss was not af-fected by turbulence even in those pipes.

For elbow and straight pipes, pressure loss in viscous flow is presumed to be different from that in turbulent flow. Pressure loss in the elbow pipe is nearly equal to that in the straight pipe at relatively small Reynolds numbers'0'. The difference in pressure loss between elbow and straight pipes becomes larger as Reynolds number increases. Beyond the critical Reynolds number, viscous flow changes to turbulent flow and pressure loss increases remarkably. This transition of airflow occurred around a friction factor of 0.036 and was independent of the radius of curvature of the elbow"'. The critical Reynolds number of elbow pipes was also reported to be larger than that for straight pipes. Measurement of air-flow in the transition zone was not carried out in this study, because only the

pressure loss within the range of the turbulent zone is needed for designing ductwork in local exhaust systems. Figure 8 indicates the relationship between coefficients of loss and the ratio of curvature radius to duct diameter for both PVC elbow and metal stamped elbow

pipes'. Coefficient of loss of the metal stamped elbow decreased with an increase in R/d, and the same tendency was also observed for PVC elbow pipes. The coefficients of loss in PVC elbow pipes were not equivalent to those of metal stamped elbow for any R/d. However, the differences in the coefficients between metal stamped elbows and PVC elbows became smaller with larger R/d values.

The coefficient of loss in the PVC elbow is presumed to be equivalent to that of the metal stamped elbow for R/d values ranging from 1.5 to 2.5. The PVC elbow

pipe has been used to join a straight VP pipe (thick-walled PVC pipe: JIS K 6741)''' to change the direction of airflow. Because the VP pipe and PVC elbow

pipe are much thicker than the VU pipe, the joint between the PVC thick-walled elbow and the straight VU pipe tended to be uneven, as compared to the joint between the elbow and the VP pipe. The uneven fitting of the joint is presumed to produce a vortex which causes increased pressure loss in the duct of the PVC

elbow and the VU pipe.


The pressure loss of PVC elbow pipes is larger than that of metal stamped elbow

pipes, but such pressure loss of the PVC elbow pipe can be reduced to greater extent than that of the metal stamped pipe by smoothing the uneven joint. Therefore, PVC material is more suitable for the ductwork of local exhaust sys-tems than metal material, because the friction loss of straight PVC pipe is far less than that of galvanized metal pipes (A PPVC = 2/3 A Pga,va)'~. It can be concluded that the use of PVC as a material for ducts effectively reduces loss of pressure and friction, leading to efficient cost-performance in construction and daily opera-tion of local exhaust systems.

ACKNOWLEDGMENTS

The authors thank Dr. Heihachiro Arito and Dr. Norihiko Kohyama of National Institute of Industrial Health, for revising the manuscript.

Fig. 8. Comparison of PVC elbow pipes with steel (stamped) elbow pipes.


REFERENCES

1) Sprenger R. Taschenbuch fur Heizung Luftung. Klimatechnik: Oldenbourg, 1974. 2) Iwasaki T, Ojima J. Friction loss in straight pipes of unplasticized polyvinyl chrolide. Ind Health

1996; 34: 245-56. 3) Inoue U. Duct Design Handbook, Maruzen 1980; 25 (in Japanese). 4) ASHRAE. Fundamentals Handbook, 1985; 33: 12-26. 5) ACGIH. Industrial Ventilation, 1995; 22nd Ed.5.36. 6) JIS. Unplasticized Polyvinyl Chloride (PVC) Pipes Fitting for Drain, JIS.K6739: 1986; 4-5 (in Japanese).

7) Tanaka K. Measurements of flow rate of air in JAKOB tube. J Aeronautical Res Inst, Tokyo Imperial Univ, 1933; 104: 213-26 (in Japanese).

8) Iwasaki T. Vena contracta and pressure loss characteristics in duct of round opening under suction. J Working Environment 1989; 10: 64-72 (in Japanese).

9) Ito H. J Jpn Soc Mech Eng., 1959; 62(490): 1634-43 (in Japanese). 10) Richter H. Forschungsheft 1930; 338. 11) White CM. Proc.Roy.Soc., Series A 1929; 123. 12) JIS. Unplasticized Polyvinyl Chloride (PVC) Pipes, JIS.K6741 1984; 1 (in Japanese).

Pressure Loss in Elbow Pipes of Unplasticized Polyvinyl ...

Documents

Transcript of Pressure Loss in Elbow Pipes of Unplasticized Polyvinyl ...