Você está na página 1de 3

Pressure Drop in the Flow of Compressible Fluids

WALTER E. LOBO, LEO FRIEND, AND G. T. SKAPERDAS The M. W. Kellogg Company, New York, N. Y.

HANGES in fluid density By means of the Fanning Pressure drop in the flow of compressible and kinetic energy are equation : fluids is affected by changes in kinetic wholly negligible in presenergy and fluid density to a degree desure drop calculations for pendent on the flow in question. An liquids flowing in conduits of integrated formula must include the effect uniform cross section. In the flow of compressible fluids, on of kinetic energy changes to be fundaand the equation the other hand, these two facmentally sound. Such a formula, which tors may affect the pressure -dF = VdpF (3) can be solved readily by means of a simple drop appreciably, particularly chart, is presented. Use of the formula for high-velocity flow a t low It follows that and chart is illustrated, and the results pressures. Complete calculation of pressure drop in this obtained are compared with those from 2 fG2v2dL dF = ___ latter case involves integration (3-4) other methods. and is rather complicated. For In the example given it was found that this reason a rapid means of Furthermore, it is obvious that the pressure drop calculated without conmaking the calculation accusidering kinetic energy changes was low rately should be useful. d b u ) = pdv vdp (4) A chart, whose use in conby 25 per cent of the correct value calcunection with the Fanning lated to include kinetic energy changes. By definition, equation avoids integration Evaluation of a dimensionless group, found for the case of compressible in the course of the derivation, can be used f l u i d s , was p r e s e n t e d b y V=Gv (5) to determine the need for considering Goodman ( I ) , but he did and for flow in conduits of not consider kinetic energy kinetic energy changes. constant cross section, changes, so that the chart is limited to those cases in d V = Gdv which kinetic enernv changes are negligible. A more (6) useful, though approximate, integrated formula described Expanding the kinetic energy expression, d(V2/2g), Equation in for below is available (4) calculation of pressure drop of com1 and solving with Equations 5 and 6, it follows that pressible fluids. Though this formula allows for kinetic energy changes, the assumptions involved in the integration make it inaccurate if there is more than a twofold variation (6-4) in density of the fluid. A chart which is not subject to these limitations is preSubstituting Equations 3A, 4, and 6 8 in Equation 1, sented in Figure 1. It is to be used in connection with the Fanning equation, as indicated in the following sample calculation which also illustrates the convenience of the chart and the extent to which pressure drop for compressible fluids is due to changes in kinetic energy. The method of constructDividing through by v2, it follows that ing the chart is developed below. At the same time the relations between the chart being presented, the chart of Goodman, and the approximately integrated equation are indicated. For flow in general, where a value of k can be found such that pvk is constant, since Construction of Chart

The starting point in calculations of pressure drop for compressible fluids is Bernoulli's equation, which for a differential section of horizontal pipe becomes:
p dv = d(pv)

kpdv

- vdp

(8)

further transformation of the equation gives


(9)
821

+ d (E) + dF

(1)

822

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
p

Vol. 34, No. 7

While Equation 15 is limited to isothermal flow for which and f are constant, Equation 10 is valid for any type of flow, as long as horizontal pipes of constant cross section are being considered. For inclined or vertical tubes Equation 1 should be modified to include a potential-head term, but this would complicate the development considerably. In the majority of cases of flow of compressible fluids, however, changes in potential head are unimportant and need not be considered. Furthermore, the expression 1 - (G2v/gkp) can never be negative because the critical velocity, which cannot be exceeded, is given (3) by
G =

$ o 3 -

r V = d s

Evaluation of the dimensionless group, G2vl/gpl, and inspection of Figure 1 will indicate the relative importance of kinetic energy changes. The approximate equation mentioned above (4) allows for changes in kinetic energy and can be derived from Equation 7, which can be rearranged to give:

4
FIGURE 1

- dp

dv f QD

2 fG2vdL

Equation 9 is similar to the Fanning Equation 2 when it is expressed as

If variations in u in the Fanning term are neglected, an approximate integration of Equation 16 can be made by using the arithmetic average of the terminal values, uav,in place of v in the Fanning term,

L '

@PA

The only difference is the addition of the expression in brackets which allows for changes in kinetic energy. Equation 9 may be put into suitable form for integration by means of the gas law equation:
pv =

M RT = B

(11)

which allows for changes in kinetic energy but is limited by the assumption of a constant average specific volume in the Fanning term. The chart presented by Goodman was based on the assumption that the change in kinetic energy could be neglected in pressure drop calculations. This assumption simplifies Equation 1 into the following equation:
pdv = d ( p v )

giving

+ dF

(18)

Integrating between p l and p z and 0 and L for the case where B is constant-that is, for isothermal flow (k = 1) in which p and f are constant:

If this equation is transformed in the same way as was Equation 1, the resulting equation obtained becomes, in place of Equation 15:

This equation can be used to calculate pressure drop. It is inconvenient, however, when the pressure drop is the unknown, because trial-and-error solution is necessary. Suppose pressure drop is defined by an equation of the Fanning type,

This equation is the basis of Goodman's chart and is included in Figure 1 as the special case for G2vl/gp, = 0.

Sample Calculation and Comparison of Methods


A hydrocarbon vapor a t 990' F. is flowing through a 4.0inch i. d. pipe a t a rate of 15,900 pounds per hour. It is required to calculate the pressure drop in 75 feet of pipe if the inlet pressure is 35 pounds per square inch absolute, and the density a t the inlet is 0.115 pound per cubic foot. The viscosity of the vapor is 1.31 X pound/(foot)(second) at 990" F. and 35 pounds per square inch absolute. 1. SOLUTION MEANS FIGURE BY OF 1.
G =

where p ~ the pseudoterminal pressure, may be considered , a n equivalent outlet pressure such that Equation 14 will give the correct result. It follows, then, from Equations 11, 13, and 14 that

(-)
P1

(-)
.'io cu. ft./lb.

50.6 lb./(sec.)(sq. ft.)

and that p s / p l is a function of p*/p1 and of the dimensionless group, G2vl/gpl or G2/gp1p1. Equation 15, which is plotted in Figure 1, can be used together with Equation 14 to calculate pressure drop conveniently,

DG _
p

50'6 = 1,290,000 12 X 1.31 X lo-' 0.0038 from Pigott's friction factor correlation (2)

v1=-=8

July, 1942

INDUSTRIAL AND ENGINEERING CHEMISTRY

823

By means of Equation 14,


Pl

- pE

2 fG'lh1

OD

since p l = 35 Ib./sq. in. abs.


p a = 35

2 X 0.0038 X (50.6)* X 75 X 8.70 X 12 32.2 X 4 = 1180 lb./sq. ft. 8.2 lb./sq. in.
e

- 8.2 = 26.8 lb./sq. in. abs.


o.137

as against 25.1 pounds per square inch absolute assumed. Hence the value of 9.9 pounds per square inch is the correct solution, which again is 22 per cent lower than the value obtained by the more accurate method. Table I summarizes the calculations. The assumption, implied in Table I, that the proposed method gives the correct result, is limited only by small variations in viscosity and friction factor.
TABLE SUMMARYOF CALCULATIONS I.
Pressure Drop Method Lb./Sq.'In. Fig. 1, with allowanoe for kinetic energy change 12.7 Integrated formula, using av. sp. vol. and allowance for kinetic energy change 14.5 Fig. 1, with no allowance for kinetic energy change 9.5 Fanning equation, using av. EP. vol. but no allowanoe for kinetic energy change 9.9

GAi gpi

(50.6)' X 8-70 32.2 X 35 X 144

@ = 0.638

P I

from Figure 1

p2 = 0.638 X 35 = 22.3 lb./sq. in. abs. AP = 35 22.3 = 12.7 lb./sq. in.

% Error
0

It is estimated that if the vapor had been flowing upward in a vertical pipe, the pressure drop would have been increased to approximately 12.8 pounds per square inch. 2. SOLUTION APPROXIMATE BY FORXULA ALLOWING FOR KINETIC ENERGY CHANGES.To use the approximate equation, trial and error is necessary to get the value of vaY. Assume p z = 20.5 pounds per square inch absolute:
v2
VI

+I4
-25

-22

'lP1

= 8.70 CU. = 8'70

-n, = 8'70 35 20.5 .ft./lb. .

= 14.84 cu. ft./lb.

US"

+ I4.g4 =11.77 cu. ft./lb. 2 G= 2 fG2v L (50.6)'(14.84 8.70) p l - p2 = - (vz - V I ) J = 32.2 s s o 2 x 0.0038(50.6)2 X 11.77 X 75 X 12 = 488 + 1600 32.2 X 4 2088 lb./sq. ft. 14.5 Ib./sq. in. p2 = 35 14.5 = 20.5 Ib./sq. in. abs.

The calculations indicate clearly that serious error can result if kinetic energy changes are always neglected. However, as the term G2vl/gp1 decreases, or p,/p1 or p ~ / p ~ increases, the importance of kinetic energy changes decreases. Thus for low-velocity flow a t high pressures, good results can be obtained with the approximate formulas, but for lw high velocity f o at low pressures, the approximate formulas are inadequate. The value of G2vl/gpl can serve as a useful index of the need for considering kinetic energy changes. The proposed method is simpler as well as more accurate than the integrated formula which uses an average specific volume, since the latter method requires trial-and-error calculations. It appears also that the errors involved in a formula integrated by approximate methods are larger when kinetic energy changes are important.

As the initial assumption is correct, the pressure drop calculation is satisfactory. The result, however, is 14 per cent
higher than that obtained by the more accurate method. Had p2 been assumed to be 22, the calculated pressure drop would have been 13.5, which would have made p2 equal to 21.5. Thus an error of 2.5 per cent in the assumed value of pa would cause an error of 7 per cent in the calculated pressure drop, making the trial-and-error calculation laborious since the result obtained is so sensitive to errors in the assumed exit pressure. 3. SOLUTION NEGLECTING KINETIC ENERGY CHANGES. A . By Goodman Chart. As above p ~ / p ,= 0.765, and p2/pl = 0.728 from Figure 1:
p2

Nomenclature
Any consistent set of units-e. g., pounds, feet, secondsmay be used: B = pRT/M d p p = pressure drop due t o friction alone, exclusive of kinetic energy effects D = diameter of tube = friction factor friction head g = acceleration due t o gravity G = mas8 velocity IC = exponent in the gas law pvk = constant for a polytropic change L = len th of pipe i = rnolfecular weight of fluid M p = absolute pressure p~ = pseudoterminal pressure, defined by Equation 14 R = gas law constant for one mole T = absolute temperature v = specific volume of fluid V = linear velocity of fluid M = compressibility factor, p v M / R T p = density of fluid
~ = l

= 0.728 X 35 = 25.5 Ib./sq. in. abs.


= 35

- 25.5 = 9 . 5 lb./sq. in. abs.

which is 25 per cent lower than the value obtained by the more accurate method.

B. By Arithmetic Average Specific Volume and Fanning Equation. Assume pz = 25.1 pounds per square inch absolute:
v1 =
VI
vav

(g) (A5)
8'70

= 12.12 cu. ft./lb.

= 8.70 cu. ft./lb.= 12'12

Subscripts 1 = inlet conditions 2 = outlet conditions av = arithmetic average of terminal values

10.41 cu. ft./lb.

Literature Cited
(1) Goodman, William, Heating, Pipino, Air Conditioning, 13, 145 (1941). (2) Pigott, R. J. S., Meoh. Ens., 55, 497 (1933). (3) Vennard, J. K., "Elementary Fluid Meohanics", p. 10, New York, John Wiley & Sons, 1940. (4) Walker, Lewis, McAdams, Gillilsnd, "Principles of Chemical Engineering", 3rd ed., p. 91, New York, McGraw-Hill Book Co., 1937.

By the Fanning equation,


Pl

- p2 = 2fG2LvSr SD
~

PO = 35

2 X 0.0038 X (50.6)' X 75 X 10.41 X 12 32.2 X 4 X 144 = 9 . 9 lb./sq. in. 9.9 = 25.1 lb./sq. in. abs.

Você também pode gostar