You are on page 1of 2

HEAT TRANSFER OF SINGLE HORIZONTAL FINNED TUBES AND

THEIR BUNDLES IN A FLUIDIZED BED OF LARGE PARTICLES

I S.S. Zabrodskyl.A.1. Tamarin, A.F. Dolidovich,


G.I. Palchonok, and Yu.G. Epanov

Luikov Heat and Mass Transfer Institute, Minsk, BSSR, USSR

INTRODUCTION

In most of the fluidized-bed boiler designs horizontal tube bundles immersed in the bed
of large particles are used for heat removal. Application of finned tubes may provide higher
heat transfer rates. But the data on finned tubes-to-fluidized bed heat transfer available up
to now pertain only to relatively small particles. The most comprehensive research work
has been done dealing with rather thin, high and close spaced fins (Petrie et ai, 1968;
Genetti et ai, 1971; Bartel and Genetti, 1973; Priebe and Genetti, 1977). The data reported
for more loosely spaced fins of low thermal resistance are contradictory from the stand-
point of the fin efficiency (Natush and Blenke, 1973; Gelperin et ai, 1972).

The present work has been intended to measure heat transfer coefficients of single
horizontal tubes with quite high and wide spaced fins as well as of their bundles immersed
in fluidized bed of large particles.

EXPERIMENTAL

The experiments have been conducted in a rectangular (25x40 cm) steel column sup-
plied with easy removable plexiglass front wall. The distributor consists of two perforated
stainless steel plates with thick cotton cloth sandwiched between them. 10 mm dia orifices are
stag ge red in the plates. The pressure drop across the distributor has been correlated by
the following equation: l:.Pd = 4x10 3 U1.35, Pa.

·S i Iica sands (d = 0.25 and 0.62 mm) , millet (d = 2.0 mm) and crushed fire clay
particles (d = 0.975,2.0,3.0 mm) were used as bed materials.

The static bed height was 30 cm in single tube runs and 40-50 cm in the runs with
tube bundles. In all the runs a single tube probe was fixed 200 mm above the gas distributor.
The probes used were 170 mm long copper tubes (finned or bare) with 30 mm outside and
10 mm inside diameters. The finned probes were turned of solid pieces of copper. The

195

J. R. Grace et al. (eds.), Fluidization


© Plenum Press, New York 1980
196 S. S. ZABRODSKY ET AL.

triangular fins were: 5 mm at base, 5 to 20 mm high and spaced 5 to 30 mm apart. The


finning factor I{! varied from 1.24 to 7.22. The probe was heated with a nichrome wire heater
located ins de the copper tUbe. 2 mm thick cork spacers and textolite pugs had been used
to minimize end losses. An orifice was drilled in the probe wall to allow chromel-copper
thermocouple 0.2 mm dia wires inside the probe and then to the potentiometer. The
thermocouple junction was welded to the outer probe surface with a silver solder.

In a special set of runs at fixed operating conditions the probe was turned about its
horizontal axis to find any wall temperature difference over the probe circumference. The
variation of the relevant potentiometer readings proved to be less than 1%.

The heat transfer coefficients were determined by the steady-state method. The
temperature difference between the outer probe surface and the bulk of the fluidized bed
was measured directly with two differentially-connected thermocouples accurate to O.5K.

In the runs with in-line and staggered tube bundles. the heat transfer coefficient was
determined by the local modelling method (Stasyalyavichus and Skrinska, 1974) with one
or two heated tubes in the array of unheated ones. Vertical and horizontal pitches in every
array were equal but varied (60; 75; 100 mm) from array to array. In the majority of the
runs the probe was placed at the centre of the third horizontal row of tubes from the air
inlet. Sometimes the probe was set in the first, second or fourth rows of the tube bundle.
The bundles themselves consisted of four or five horizontal tube rows with the bottom
(first) row fixed 150 mm above the distributor.

The effective heat transfer coefficient of afinned tube, based on the bare tube surface,
was determ ined as

(1)

The heat transfer coefficient for a finned tube based on its total surface area was calculated
by the same equation replacing Ab by Aft-

In order to estimate the probe end heat losses a special procedure has been applied. The
whole probe (with the heater removed) preheated in a furnace up to 1000 C was then cooled
in the fluidized bed, and the heat transfer coefficient without end losses was determined
by the well-known regular regime method (Kondratiev, 1954). The comparison between
the heat transfer coefficients measured under the same conditions by the above steady-
state and regular regime methods allows evaluation of the end losses in the experiments
performed. These end losses are found to range from 5 to 10% and are taken into account
when handling experimental data.

EXPERIMENTAL RESULTS AND DISCUSSION

Single Finned Tube

Representative heat transfer coefficient - superficial gas velocity relationships, «tb =

You might also like