Professional Documents
Culture Documents
Correlacion de CAIDA DE PRESION en Lechos Empacadosarticle PDF
Correlacion de CAIDA DE PRESION en Lechos Empacadosarticle PDF
Correlacion de CAIDA DE PRESION en Lechos Empacadosarticle PDF
Abstract
This work aims at validating and extending the applicability of an existing correlating equation for predicting pressure drop through packed
beds of spheres. A new set of data points out its use for bed geometric aspect ratio D/dpart in the range 3.814.5, in the case of dense packings.
The case of loose packings is also discussed using literature data. As a conclusion, a correlation is proposed for predicting pressure drops through
fixed beds of spheres (dense and loose packings) for large ranges of Reynolds numbers (10 < Repart < 2500) and of geometric aspect ratio
(3.5 < D/dpart < 4050).
2006 Elsevier B.V. All rights reserved.
Keywords: Packed beds; Spherical particles; Laminar flow; Turbulent flow; Voidage; Geometric aspect ratio
0255-2701/$ see front matter 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.cep.2006.07.002
330 A. Montillet et al. / Chemical Engineering and Processing 46 (2007) 329333
Table 1
Main literature equations proposed for pressure drop predictions in large ranges of Reynolds number
Author/reference Equation Range of validity in terms of Repart
Presently, these equations are presented in terms of a modified particle friction factor fpart vs. the Repart .
in the beds in order to ensure sufficient accuracy in the deter- the aqueous solutions of glycerol is measured from a sample of
mination of the bed porosity. The uncertainties brought about liquid taken from the experimental apparatus and with a Couette
by the use of flow-meters and manometers must be minimised viscosimeter (Reference: Rheoanalyzer, Contraves, module: MS
since a great accuracy is needed for the data in view of the sub- 145). Commercial particles have been used such as glass par-
tle scatter observed between some sets of experimental pressure ticles or particles made of a rigid synthetic material. All the
drop variations and a quadratic Forchheimer type equation. used particles are smooth and rigid. Tables 2 and 3 recapitulate
The objective of this work is to present a new set of exper- the main characteristics of the tested beds, particles and experi-
imental data in order to discuss and confirm the usefulness of ments.
the previously proposed correlating equation. In conclusion, the
relevance of the proposed equations is discussed, based on a 3. Experimental results and procedure for correlating
comparison with correlating equations from the literature. the data
Table 2
Characteristics of new data (a.q.s.: aqueous solution of glycerol)
System Characteristics of the fluids Data
Particles, dpart (mm) Geometric aspect Fluid Density at 20 C Dynamic viscosity at Range of Reynolds No. of data
ratio, D/dpart (kg m3 ) 20 C (mPa s) number, Repart points
Table 3
Main bed and particle characteristics
System (particles/bed) Range of Repart (Eqs. (8) and (9)) MRS (%) and number of data
Mean relative scatter (MRS) between computed data (Eqs. (8) and (9)) and new experimental data.
Table 4
Main characteristics of some loose packings tested by Griffiths [14]
System Range of Repart (Eq. (8)) Relative scatter (%) between (Eq. (8) with
a = 0.050 and b = 0.20) and experimental
dpart (mm) D/dpart
Relative scatter between the values of fitted to experimental correlating equations and the computed values of (Eq. (8) with a=0.050 and b=0.20).
measurement data. Most of them are easily usable and cover a quantify. Zou and Yu [13] have observed that for D/dpart > 1.8,
relatively large range of Repart , namely 301000. the variation of the bed porosity as a function of D/dpart follows
These data have been analyzed using the procedure of com- two different paths, for loose packings and dense packings.
putation presented in Section 3.1. Eq. (8) is still convenient for
the representation of these data, using b = 0.20, but a new value 4. Discussion and conclusion
for the constant a is needed, yielding: a = 0.050.
The values of obtained by a fitting with the experimental 4.1. Fields of applications of Eq. (8) for prediction of
equations and those obtained from the proposed correlation for pressure drops
are compared in Table 4.
The dense or loose nature of the packing does not influence The form of Eq. (8) is a power law. The upper limit of its use
the power of the geometric aspect ratio in Eq. (8), but the value in terms of D/dpart is 40 which is not a real problem as the wall
of the constant a depends on the nature of the bed. It means that effects are often considered to be negligible for D/dpart > 4050
the particle organisation, is not yet properly taken into account [3]. If this assumption is admitted, a limit could be fixed for
in the proposed equation. This parameter is rather difficult to D/dpart > 50: (D/dpart )0.2 2.2. The case of beds with aspect
ratios D/dpart < 3.5 is special and should be considered sepa-
rately. In particular, our equation does not represent accurately
the data of Calis et al. [15] obtained for D/dpart 4. Griffiths [14]
noticed that the porosity variation is reversed for D/dpart < 3.5
and the data of Zou and Yu [13] confirm this phenomenon.
31% is calculated. The present correlating equation seems to fluid viscosity (Pa s)
increase noticeably the reliability of prediction of experimental fluid density (kg m3 ).
data, especially for the tested narrow beds.
References
4.3. Conclusion
[1] A. Montillet, Flow through a finite packed bed of spheres: a note on the
The correlating equation that we propose for predicting pres- limit of applicability of the Forchheimer-type equation, J. Fluids Eng. 126
(2004) 139143.
sure drops through beds of spheres and the corresponding ranges [2] H.E. Rose, On the resistance coefficientReynolds number relationship for
of applicability are the following fluid flow through beds of granular materials, Proc. Inst. Mech. Eng. 153
(1945) 154161.
1 D 0.20
fpart = a [1000 Re1 0.5
part + 60 Repart + 12]
[3] H.E. Rose, A.M.A. Rizk, Further researches in fluid flow through beds of
3 dpart granular material, Proc. Inst. Mech. Eng. 160 (1949) 493503.
[4] H. Watanabe, Drag coefficient and voidage function on fluid flow
with a = 0.061 for dense packings or a = 0.050 for loose pack- through granular packed beds, Int. J. Eng. Fluid Mech. 2 (1) (1989) 93
ings. 3.8 D/dpart 4050 and 10 < Repart < 2500. 108.
For D/dpart > 50, knowing that the influence of the bed geo- [5] R.E. Hicks, Pressure drop in packed beds of spheres, Ind. Eng. Chem. Fund.
9 (3) (1970) 500502.
metric aspect ratio is considered to be negligible, the term
[6] J.A. Tallmadge, Packed bed pressure dropan extension to higher
(D/dpart )0.20 can then be estimated as being equal to 2.2. Reynolds numbers, AIChE J. 16 (6) (1970) 10921093.
[7] J. Lee, K. Ogawa, Pressure drop through packed bed, J. Chem. Eng. Jpn.
Appendix A. Nomenclature 27 (5) (1974) 691693.
[8] D. Ciceron, J. Comiti, R.P. Chhabra, Pressure drops for purely viscous non-
Newtonian fluid flow through beds packed with mixed-size spheres, Chem.
a constant Eng. Commun. 189 (10) (2002) 14031414.
A constant (Pa m2 ) [9] A. Montillet, L. Le Coq, Characteristics of fixed-beds packed with
b constant anisotropic particlesuse of image analysis, Powder Technol. 121 (2001)
B constant (Pa m3 ) 138148.
[10] S.S. Ergun, Fluid flow through packed columns, Chem. Eng. Prog. 48
dpart particle diameter (m)
(1952) 8994.
D bed diameter (m) [11] I. Wahyudi, Ecoulement dans les sols granulaires satures en regimes per-
fpart modified particle friction factor, fpart = manent et transitoires, Doctorat, Universite de Nantes, Ecole Centrale de
(|P|/H)(dpart /U02 ) Nantes, 1998.
MRS mean relative scatter, MRS = [12] R.M. Fand, B.Y.K. Kim, A.C.C. Lam, R.T. Phan, Resistance to the flow
(1/n) ni=1 |((experimental value of fluids through simple and complex porous media whose matrices are
composed of randomly packed spheres, J. Fluids Eng. 109 (1978) 268
computed value)/experimental value)| 274.
|P|/H pressure gradient (Pa m1 ) [13] R.P. Zou, A.B. Yu, The packing of spheres in a cylindrical container: the
Repart particle Reynolds number, Repart = U0 dpart 1 thickness effect, Chem. Eng. Sci. 50 (9) (1995) 15051507.
U0 superficial velocity (m s1 ) [14] N.B. Griffiths, The flow in and structure of narrow packed beds, PhD
Report, Sydney Sussex College, Cambridge, UK, 1986.
[15] H.P.A. Calis, J. Nijenhuis, B.C. Paikert, F.M. Dautzenberg, C.M. Van Den
Greek symbols Bleek, CFD modelling and experimental validation of pressure drop and
coefficient flow profile in a novel structured catalytic reactor packing, Chem. Eng. Sci.
bed fractional voidage 56 (2001) 17131720.