Evaluation of prediction models on frictional pressure drop for condensation flow in horizontal circular tubes

ABSTRACT

There are numerous correlations available for predicting frictional pressure drop of condensing flow in horizontal tubes. It is necessary to evaluate and analyze the prediction accuracy and applicability of existing correlations. In order to explore the versatility of frictional pressure drop correlations for condensation flow with different working fluids, four existing models are investigated: phase-separation conversion model, full-phase flow model, two-phase multiplier flow model, and homogeneous flow model. Combined with five different working fluids (R170, R12, R600a, R22, R1234yf), 402 groups of experimental data on frictional pressure drop for condensation flow in horizontal circular tubes are considered with respect to flow pattern. The prediction results based on each correlation are compared and analyzed. Mean absolute relative deviation (MARD) and mean relative deviation (MRD) are selected as indicators to evaluate each correlation. The results showed that the versatility of the existing correlations needs improvement. The prediction accuracy of frictional pressure drop for condensation flow can be improved by selecting different correlations for different flow patterns. The phase-separation conversion model has a good prediction effect and a wide application scope in annular flow. The maximum MARD is 15.85%, and the maximum MRD is 8.72%. 93.08% of the data predicted deviation is less than 30%. This study will provide some constructive instructions to predict frictional pressure drop of condensation flow in horizontal circular tubes for designing heat exchangers.

KEYWORDS:

• Two-phase flow
• condensation
• flow pattern
• frictional pressure drop
• correlation

Nomenclature list

$\frac{\mathrm{d}p}{\mathrm{d}l}$	=	frictional pressure drop, Pa/m
tp	=	two phase
l	=	liquid
lo	=	liquid only
v	=	vapor
vo	=	vapor only
${\varphi }_{\hspace{0.17em}}^2$	=	correction coefficient
Rel	=	Reynolds number
m	=	mass flux, kg/(m2s)
x	=	dryness
ρ	=	density, kg/m3
Xtt	=	Lockhart–Martinelli parameter
μ	=	dynamic viscosity, Pa s
d	=	hydraulic diameter, m
$\mathit{\sigma }$	=	surface tension coefficient, N/m
We	=	Weber number
Fr	=	Froude number
f	=	friction factor
${\eta }_n$	=	percentage of points with prediction deviation within ±n
Abbreviations	=
MARD	=	mean absolute relative deviation
MRD	=	mean relative deviation

Disclosure statement

The authors have declared no conflicts of interest.

Additional information

Funding

The work was supported by the National Natural Science Foundation of China [No. 51906036, U21B2087]