~#717 7

ELSEVIER

Impinging Radial and Inline Jets: A Comparison

with Regard to Heat Transfer, Wall Pressure
Distribution, and Pressure Loss
F. Peper • Heat transfer and wall pressure distribution on a plane surface generated
W. Leiner by single impinging inline or radial jets are studied experimentally. The
pressure drop of inline and radial jet nozzles is measured. The effects of
M. Fiebig
flow exit angle, nozzle to surface distance, and exit velocity on heat
Lehrstuhl fiir Wiirme- und StoffiJbertragung,
lnstitut fiir Thermo- und Fluiddynamik, transfer, wall pressure distribution, and pressure drop are discussed. Heat
Ruhr-Unh~ersitiit Bochum, Bochum, transfer results show that radial jets with flow exit angles of +45 °- +60 °
Germany generate up to 60% higher local and up to 50% higher global Nusselt
numbers compared with inline jets of the same volumetric flow rate and
exit velocity. Measured wall pressure distributions are presented in terms
of pressure coefficients. The total force exerted by radial jets on a plane
surface is lower than that exerted by inline jets. Radial jets with negative
flow exit angles can generate small lifting forces. Results of pressure drop
measurements are presented in terms of resistance coefficients, which allow
an estimation of the necessary additional fan power if radial jet nozzles
instead of inline jet nozzles are employed. For radial jet nozzles with flow
exit angles of +45 ° < 0 < +60 ° the rise of fan energy costs is negligible
compared with the rise of heat/mass transfer. Radial jet nozzles have a high
potential for application, particularly when very high drying rates or small
jet forces on the impingement surface or both are required. © Elsevier
Science Inc., 1997

Keywords: single impinging radial jets, single impinging inline jets,

impingement heat transfer, impingment pressure distribution, nozzle
pressure loss

INTRODUCTION constant mass flow rate and nozzle exit velocity [2, 3].
However, slot nozzles are preferred if a very uniform
Impinging jets are applied in chemical and thermal engi- heat/tnass transfer is desired. A n inline jet produces posi-
neering to increase heat and mass transfer. Applications tive, repulsing forces on the impingement surface (Fig. 1).
include the drying of p a p e r and textiles, the tempering of A n y nozzle or orifice with a circular cross section pro-
glass sheets, and the cooling of gas turbine components or
ducing an inline jet can be converted into a radial jet (RJ)
electronic equipment. The inline jet (ILJ) nozzle or ori-
nozzle by mounting a flow guiding plate or cone trans-
fice, in which the jet axis is aligned with the nozzle's
versely in front of the exit. This modification transforms
centerline, is a typical device for generating impinging
jets. Inline jet nozzles with circular of rectangular cross the exit into a circumferential slot at the end of the nozzle
sections (slot nozzles) are in wide use. and forces the jet to leave the nozzle in a radial direction
The free jet leaving the nozzle impinges perpendicularly (Fig. 2). Fluid is entrained from the space between the
on a plane surface, where it increases heat or mass radial free jet and the impingement surface. If the vertical
transfer. Very high values of transfer coefficients right at nozzle to surface distance is small enough, this reduces
or near the central stagnation point are typical for inline the static pressure below the flow guiding plate. The radial
jets; however, the heat or mass transfer decreases rapidly jet is bent toward the surface and forced to reattach in an
with increasing radial distance [1]. Inline jet nozzles with annular stagnation line. High transfer coefficients are
circular cross sections are r e p o r t e d to produce an 8% to p r o d u c e d in the vicinity of this " r e a t t a c h m e n t ring" [4].
25% higher global heat transfer than do slot nozzles at Hence the area of high heat or mass transfer is larger

Address correspondence to F. Peper, Ruhr-Universit~it Bochum, Lehrstuhl fiir W~irme- und Stoff'tibertragung, Geb~iude IB 6/127, Universit~its-
stra be 150, 44780 Bochum, Germany.
Experimental Thermal and Fluid Science 1997; 14:194 204
© Elsevier Science Inc., 1997 0894-1777/97/$17.00
655 Avenue of the Americas, New York, NY 10010 PII S0894-1777(96)00066-0
 Impinging Radial and Inline Jets 195

a)

Figure 1. Sketch of inline jet (ILl) nozzle and impinging flow.

than that for an inline jet. The flow field of a RJ nozzle

and its essential geometric parameters are shown b)
schematically in Fig. 2.
The flow exit angle of the radial jet can be modified by
the shape of the flow guiding front plate. A radial jet flow
leaving the nozzle with a velocity component away from
the surface has a negative flow exit angle (0 < 0 °) by
definition. The force on the impingement surface can be
controlled by altering the flow exit angle. It is possible to
generate a net positive (repulsing) or negative (suction)
force on the surface. Consequently, the force can also be
reduced to zero, thereby providing a very gentle drying or
cooling/heating performance for stress-sensitive products.
Inline jets have been studied many times in the past
decades. The most important questions for the engineer in
industry--What are the optimal geometry and position of
ILJs for obtaining maximum global heat or mass transfer?
• rR I
t
- - h a v e been answered and verified by many investigators.
Some excellent reviews are provided by Viskanta [3] and Figure 2. (a) Sketch of radial jet (RJ) nozzle and flow. (b)
Martin [5, 6]. Characteristic variables for single RJ nozzle (0 = 0°).
In comparison with the numerous studies of inline jets,
only a few investigations have been made on radial jets. Available publications on radial jets are lacking in three
Page et al. [4] developed a model theory for the RJ flow important aspects: first, the vertical nozzle to surface
fields for laminar and turbulent cases. They deduced and distance (nozzle height) and, in particular, the flow exit
confirmed experimentally that, in the turbulent case, wall angle providing optimal h e a t / m a s s transfer has not been
pressure distribution, Cp(r), reattachment angle, OR, and intensively investigated yet. Second, the question of addi-
reattachment radius r R, are independent of the jet tional fan power required for feeding RJ nozzles instead
Reynolds number, Re. A dimensionless force coefficient, of ILl nozzles has not been treated in the literature. Even
C F, has been defined [7] to describe the net surface force if RJs dry or cool better than ILls, it is still important to
of radial jet flows. know the price that has to be paid for applications. Third,
Infrared measurements have shown that, depending on studies on the interaction of radial jets in periodic arrays
the area of influence considered, a single RJ with a as required in a drying apparatus are still lacking.
positive flow exit angle 0 > 0 ° can create a higher average The first and second aspects motivate the present study.
heat transfer coefficient, am, then can an 113 [8]. The Measurements of local heat transfer coefficients were
mass flow rate and the exit velocity were kept equal for conducted with the use of infrared thermography (IRT).
comparing both nozzle designs. Page [9] measured drying The wall pressure distribution at the impingement surface
performances of both an ILl nozzle and a RJ nozzle. The and the pressure loss caused by the different front plates
Page group investigated four different values of flow exit also were measured.
angles (0 = - 1 0 °, 0 °, +10, and +45°). Nozzle front plates of nine different angles (0p~ =
Recently, RJ blow boxes with flow exit angles of 0 = -60 °- + 60 °) producing actual flow exit angles of - 1 8 ° <
- 1 0 ° have been installed and tested in paper machines 0 < +70 ° [15] were investigated. Different nozzle heights,
[10, 11]. Results show that, besides the additional drying H, were tested to identify optimal RJ geometries for
effect, the sheet stability was essentially increased owing different applications (i.e., high heat transfer or low sur-
to the suction forces generated by the RJ nozzles. face force or both). All measurements were performed
Numerical studies of flow field and heat transfer by a equally for inline jet nozzles and compared with ILl
radial jet have been performed for laminar flow by literature where available. The results provide an orienta-
Laschefski et al. [12, 13]. Simulations of the turbulent tion for further investigations of arrays of radial jet noz-
nozzle flow are still in progress [14]. zles.
196 F. Peper et al.

EXPERIMENTAL APPARATUS shielding also serves to prevent forced convection effects

on the bottom side of the inconel foil. The upper side of
Test Facility the inconel foil is not painted; its emissivity is estimated to
be e u = 0.2.
A sketch of the test facility is given in Fig. 3. Ambient air A Plexiglas surface with a built-in two-dimensional
is pumped by an adjustable blower (8) into a plenum traversing system was used as an impingement surface
chamber (7). The temperature of the airflow is controlled instead of the inconel foil for measurements of the wall
by a heat exchanger that is fed by a thermostat (10). A pressure distribution. Thirteen static pressure ports are
calibrated orifice gauge (9) on the suction side of the fixed in the traversing system. The pressure signals of a
blower is used to measure the volumetric flow rate. Nine Baratron pressure meter are transmitted to a Schlum-
valves (11) are fixed at the plenum chamber, feeding nine berger isolated measurement pod (IMP) 35951B. Simulta-
nozzles in a 3 × 3 square array configuration. The present neously, the IMP is connected to a R O S E M O U N T pres-
work focuses on the study of single nozzles; hence eight sure meter recording the flow pressure drop at the orifice
valves remain closed. The nozzle (1) is fixed in a three- gauge and to three Ptl00 thermocouples measuring the
dimensionally adjustable support (2). temperature of the ambient air, T~, and the temperatures
Inconel foil is used as the impingement surface (3) for of the airflow at the orifice, Tv, or, and the nozzle exit,
heat transfer measurements. It is tightly stretched in a Tv E. The IMP is connected to a personal computer (12)
special, electrically insulated holder, which is fixed in a collecting and analyzing all data.
horizontal wooden board. The foil is connected by copper
bars to a dc power supply with built-in voltmeter (13). All Nozzles
heat transfer experiments were conducted with cold air
Aluminum nozzles with an outer radius of r o = 20 mm,
(Tv, E = T~) impinging on the heated foil.
and an inner radius of r i = 14 mm were employed. Figure
An infrared camera (5) is positioned beneath the in-
4 shows a sketch of the different nozzle shapes. To allow
conel foil, looking vertically upward onto the rear side of flow acceleration, the nozzle was conically shaped in the
the impingement area. Owing to the small thickness of the case of 0pl > 0 °, as shown in Fig. 4b. For plates with
foil of SFoil 0.05 mm and its thermal conductivity of
=
0pl _< 0 ° nonconical nozzles (Fig. 4c) were used to achieve
Avoi~ = 10.5 W / ( m K), the temperature difference be- negative flow exit angles, 0. The flow exit angle, 0, is not
tween upper and lower surface is negligible. The key identical with the front plate angle, 0pl [15], as shown in
advantage of having the nozzle and camera on opposite Figure 5.
sides of the foil is the easy optical access to the impinge- Inline jet and radial jet nozzles were compared, keeping
ment area, even for narrow nozzle arrays in which the the volumetric flow rate, t?v, equal for both nozzle types.
effect of the center nozzle has to be investigated. Only the velocity component perpendicular to any cross
The lower side of the inconel foil and all sides of the section, v x = Cv, t~ cos 0, contributes to the calculation of
wooden shielding (4) are coated with a special paint of a VF (see Fig. 5). It follows that
known homogeneous emissivity of e 1 = 0.95 to allow ac-
curate evaluation of radiation losses. The infrared camera [TF, ILJ = I/IF, RJ ¢~ UV, N o ' n ' r i 2 = L'F, E C O S 027rbro (1)
was calibrated for this special emissivity. The wooden The cross-section area and consequently the average exit
velocity were kept equal for ILJs and RJs to avoid distor-
tion of the results by different flow momentum:

UF, n o = L'F, E (2)

Substituting Eq. (2) in Eq. (1) yields the radial jet nozzle
width, b, as a function of the flow exit angle, 0, as shown
in Fig. 5:
ri2 b*
b - - - - (3)
2 r o cos 0 cos 0

a) ILJ b) R J; Opt> 0 ° c) R J; Opl< 0 °

I

ro i
Figure 3. Schematic of test facility (setup for heat transfer
measurements with infrared thermography); 1, RJ nozzle; 2,
support; 3, exchangeable impingement surface (inconel foil,
stretched in a frame); 4, wooden shielding; 5, infrared cam-
era; 6, infrared processing system AVIA TVS 2000; 7, plenum
chamber with flow straightener; 8, blower; 9, orifice gauge; Figure 4. Sketch of employed nozzle shapes: (a) ILl nozzle,
10, heat exchanger fed by thermostat; 11, valves; 12, data (b) RJ nozzle with 0pj > 0 ° (conical ILJ nozzle), (c) RJ nozzle
processing PC; 13, power supply with built-in voltmeter. with 0pl < 0 ° (nonconical ILJ nozzle).
 Impinging Radial and Inline Jets 197

O f ' c o n v ,(~ r ' o c ~ , u

u

I i~-I:'r'conv,u~1~1
, L

@] 0 1 = [ ' ~ 4
r'ad,I Cr-convAI
- J
Figure 5. Variation of nozzle slot width, b, with flow exit
angle, 0 to maintain the condition re. E = VF, No"

Only for the "ideal" radial jet with a flow exit angle 0 = 0 °
is the effective slot width, b*, equal to the actual slot
width, b. For investigations of flow exit angle 0 ~ 0 ° the Figure 6. Schematic plot of foil with control volume for
actual slot width, b, has to be changed to maintain the balancing heat fluxes: 1, control volume; 2, inconel foil with
condition of equal average exit velocities; see Eq. (2). insulated holder (e u = 0.2; e I = 0.95); 3, wooden board; 4,
Radial jet and inline jet nozzles are not geometrically wooden shielding (6 = 0.95); 5, infrared camera.
similar, which presents problems when comparing both
designs. The characteristic length of inline jet nozzles is
the inner nozzle diameter d i = 2ri; whereas, for radial jet there will be a small amount of free convection, Qfrconv,l,
nozzles, the characteristic length is the hydraulic diame- at the foil edges. The heat losses by free convection from
ter, d h = 2b*. For all nozzles employed here, the constant the upper and lower sides are measured together after a
ratio between both characteristic lengths is steady state has been reaches ( = 30 min) by electric
heating with a disconnected fan.
d i = 2.857(2b*) ~ r i = 2.857b* (4)
. . . . = Q f r . . . . . . ÷ Q f r ..... , (7)
However, to simplify the comparison, the Reynolds num-
ber, Re [and consequently the Nusselt number N u - - s e e From Eqs. (6) and (7), the specific local convective heat
Eq. (11)], is defined in the same way, with the inner nozzle flux, qfcon.... is calculated as
d i a m e t e r , di, for inline and radial jet nozzles: EI
t]f . . . . . . AFoi I t~fr. . . . -- ¢)rad, u -- t)rad,l' (8)
UF,Nodi
Re (5)
P F, No The local heat transfer coefficient, a, is calculated from
Eq. (8) as
The uncertainty of determining the Reynolds number was
calculated according to the standard Gaussian method qfgonv, u
ct (9)
[16] to be less than 4.5%. TFoil- TF, E "
Average heat transfer coefficients, am, are calculated for
HEAT TRANSFER EXPERIMENTS every circular area, ~-r~2, with 0 _< r n < 10 r i by averaging
Analysis ~)fc. . . . and TFoil o v e r a corresponding number of n circu-
lar ri'ng areas, with an absolute ring width of Ar = 1.45
A sketch of the control volume enclosing the inconel foil, mm:
with different heat fluxes entering or leaving it, is given in
Fig. 6. The dc-power supply generated a total electric heat 1/qy"n= 1 tTfco. . . . . i(r/2 -- ri2-1 )
(10)
flux, Qel, supplied to the foil. Different heat fluxes leave Olin: 1/r;,2~i=n 1 TFoil,i(r/2 -- r/2-1 ) -- TF, E
the foil: on the upper side, the heat fluxes by forced
The local and average Nusselt numbers, Nu and N u m ,
convection, Qfco..... free convection, Ofrco . . . . . and radia-
respectively, are defined as
tion, Qrad, u, respectively; on the lower side, the heat fluxes
by radiation, Qrad, 1, .and free convection Qfrconv, 1. The ad i
tangential heat flux, Qcond, by conduction from the inconel Nu Nufr . . . . . (11)
AF, No
foil to the holder is neglibible owing to low thermal
conductivity. A steady state energy balance yields o~md i
Num Nufr . . . . (12)
AF, No
(~el = E 1 = O f . . . . . ÷ Off . . . . . .
Here Nufrco,v is the free convection Nusselt number for
the equivalent average temperature difference, Tfoil-
÷ 0 r a d , u ÷ 0 r a d , , ÷ Off . . . . . 1 (6)
TF, E; see Eq. (14).
The air layer beneath the heated foil will essentially It is common to use the adiabatic wall temperature, Tad,
remain stable and heat up very slowly. However, because, instead of the jet exit temperature, TF, E, in Eqs. (9) and
the wooden board, unlike the inconel foil, is not heated, (10) to compensate for compressibility effects of the jet
198 F. Peper et al.

flow in the stagnation point regions. This effect has been Inline Jet Heat Transfer The local Nusselt number,
neglected in the evaluation of the infrared measurements NUlo~, is plotted in Fig. 8 vs the dimensionless radius, r / r i ,
described here owing to the relatively small flow exit for inline jets with different relative nozzle heights, 1 <
velocity VF. ~ = 30 m / s = constant, which is c o m m o n for H / d i < 8. The measured data agree well with results of
many industrial drying applications. The resulting Mach other investigators [19, 20] in predicting a first and second
number is Ma = 0.087, from which the adiabatic wall annular peak of heat transfer outside the centerline for
temperature at the stagnation point, T~d,st can be approxi- small nozzle heights. The outer peak is due to a transition
mated [17]: from laminar to turbulent flow in the boundary layer. The
reason for the inner and less-pronounced peak is still
T"dSt
• ,, = TF, E( 1 +3 P ~ K - "~ 1 Ma 2 ) _< 1.0013TF, E (13) under discussion. A suggestion is that the inner peak is
due to a higher turbulence level at the circumference
Here K is the isentropic exponent of air. Equation (13) (r = r i) of the jet than in the center [20]. It is supposed
shows that the error made by neglecting the difference that tangential flow from the central stagnation area of
between TF, ~ and T,~a,st amounts to <0.4 K. It disap- the jet causes this rise in turbulence level. With increasing
pears at small distances from the stagnation area. nozzle to surface distance, the outer peaks disappear and
the inline jet heat transfer distribution has its well-known
bell-shaped curve with a maximum at the jet's centerline.
RESULTS
Optimal local heat transfer is achieved at a dimensionless
Influence of Free Convection The Nusselt number, height of H / d i = 7, which is similar to the assumed
Nufr .... = C~tr.... di/a(T-~) caused by free convection on length of the jet's potential core. For further increasing
both sides of the inconel foil, [Eq. (6)] was determined for distances, local heat transfer at the impingement point
different temperature differences, Tfo. - T~, as shown in decreases owing to the jet's loss of momentum.
Fig. 7. The measured data were fitted by the correlation The average Nusselt number, Num, is plotted in Fig. 9
vs the reciprocal scaled nozzle area, f i. The reciprocal
Nuf ..... = 1.403(Tfoil- T~) 0"435. (14) relative nozzle area, f - 1 , is defined as the cross section of
the nozzle divided by the circular area of influence at-
A correlation for turbulent free convection on top of a
tached to it:
heated surface by Stewartson [18] is presented in Fig. 7 for
comparison:
Acircle (r) 2
Nu frc,,nv,u - di
1 0.15[Gr Prf2(Pr)] °33 (15) l/f= AN,, -- 77 . (17)

with
A correlation given by Schiinder and Gnielinski [1] valid
f2(Pr) = [1 + (0.322/Pr) 11/2°] 20/11 (16) for 5 _ < r / r i < 15, 2 _ < H / d i<_ 12 and 3 × 1 0 4 < R e _ <
1.2 × 105 is plotted in Fig. 9 for comparison:
Figure 7 shows that the measured data exceed the values
predicted by Stewartson's correlation. The difference is Nn m di 1 - 1.1di/r
probably caused by free convection effects near the edges 0.54Re 2/3. (18)
on the foil underside. These effects seem to vanish for Pr °'42 r 1 + O.l(H/d i - 6)di/r
small temperature differences between foil and ambient
air but increase with higher temperature differences. For It becomes evident that inline jets with maximum local
the calculation of the forced convection local and global heat transfer ( H / d i = 7) produce high average heat
Nusselt numbers [Eqs. (11) and (12)[, the free convection transfer only within a small area of f 1 < 10. If high
Nusselt number, Nufrconv, for the equivalent average tem- average heat transfer is to be achieved across a larger
perature difference, Tf,,~ - Tv, E, is determined from Eq. area, it is more benefitial to reduce the distance between
(14) and subtracted. the ILJ nozzle and the impingement surface ( H / d i ~ 3).

Influence of free convection

5 ........... .~ . . . . . . . . . . . ', . . . . . . . . . . . . : ........... : ............ ; ........... "} ........... :"'"'~'~

i i i i
_ ~ - . . . . . T- ° i
4 ........... + ........... .:............ , ........;-~} -'-....... !........... *........... !............
=-
O
e
z

Figure 7. Free convection Nusselt number, Nufrconv, plot-

ted against temperature difference between foil and ambi-
ent air, Tfoil- T~. Solid line: correlation for turbulent free
1

03
.
1:2:iZ22i12::I
.
5
.
7
--~-- data; Nu = 1 . 4 0 3 AT o . 4 ~

.
9
correlation;J1 81

11
i
13
,
15
, i
17
I::::I
19
convection on top of a heated surface [18]. Tfoil - T . . [K]
 Impinging Radial and lnline Jets 199

a)
220 in-line iot : = ' i I 400 r : : :
2oo p
.- .
Re = 54000"
,
. .
. . .
:
VF E = 3 0 r r l / . q
. .
:
.
:
I o
[]
Hid=, , 1
r i I~L i
Radial jet; 0 = +55 °
Re = 5 4 0 0 0 ; w'.E = 3 0 m / s

~=.~. ......... ~........... L . ~ ........ "-'-"-I & H/di = 5

3oo .............. .............. t .............. : .............. . . . . . . . . . . . . . . .

180 ~ i ~ i I 0 Hldl=7
i= i
-i ~ i |=R
z1 1,01 iiiiiiii
o .........i...........
1 2 0 ........... i ........... ~........... i........... I. . . . . . . . . . . . .
Z 200 ~

l°°
"

io....... T..............
i
1 O00 2 3 4 5 6 7 8 0 4 8 12 16 20 24
r/n r/b*

Figure 8. Local Nusselt number, Nu, of inline jet flow plotted

against dimensionless radius r/r i for different scaled nozzle
b) 400
heights H/di; Re = 54,000.
Radial jet;0"= +40
Re = 54000; VF.E ----- 3 0 m / s

The improvement in global heat transfer is due to the 3O0

i.......~ .............T i ..........".............. ~............i
outer peaks in local Nusselt number distribution, which ! 0 .% i l 0 H//b*=2.041
i o i~ i I 0 Hlb. = 4.08 1
result in high local heat transfer over an annular area, 7~ 200 .............. ii.............
O ' - ~]'.........
~-i ~"" OA H/b*H~b*
==8.166"12 [.
similar to radial jets. The measured data differ from the •

correlation by Schliinder and Gnielinski [1] by less than

7%.

Radial Jet Heat Transfer The influence of nozzle height

on local heat transfer by RJ nozzles with effective flow 0 ' 4 ' 8 ' 12 ' 16 ' 20 ' 24
exit angles of 0 = 55 ° and 0 = +4 ° is shown in Fig. 10. r/b"
Results generally show that, for radial jet nozzles, maxi-
mum local heat transfer is achieved just before the reat- Figure I0. Local Nusselt number, Nu, of radial jet flow
tachment ring. This conclusion can be drawn because the plotted against dimensionless radius r/b* for different scaled
radial location with maximum local Nusselt number is nozzle heights H/b*; Re = 54,000. (a) 0 = +55°; (b) 0 =
+4 °.
always slightly smaller than the radius with maximum wall
pressure, as will be shown below. The reattachment ra-
dius, r R, increases with increasing scaled nozzle height,
H/b*, and the local maxima of Nusselt number or heat H / b * = 2.04 (0 = +4°). Moreover, the +55 ° nozzle
transfer coefficient are up to 60% higher compared with achieves higher local heat transfer coefficients in the
an inline jet at H / d i = 7. central area right beneath the flow guiding plate r/b* <
In comparing Figs. 10a and 10b, it further becomes 4.08.
evident that, despite the large difference in flow exit The influence of flow exit angle, 0, at constant scaled
angles, the peak local Nusselt numbers differ by less than height, H/b*, on local heat transfer is shown in Fig. 11.
15%. However, optimal values are achieved at different Figure 11 shows that the reattachment radius, r R, in-
relative nozzle heights H / b * = 4.08 ( 0 = +55 °) and creases with decreasing flow exit angle, 0. Further, it
becomes evident from Fig. l l a that, at a scaled nozzle
height H/b* = 2.04, radial jets with extremely positive
flow exit angles (0 = + 70 °, +55 °) achieve poorer maxi-
220 In-line'jet ' ! ! ! I
L.Re = 54000; VF,E = 30 m / s .. 0 H/dl = 1
mum local heat transfer than do RJs with smaller values
2 0 0 ~ :: i i i i i I 1:3 H/di=3 of 0 ( - 5 ° < 0 < +18°). However, if the scaled nozzle
,~ i i :: i :: i / A H//d, = 5 height is increased to H/b* = 4.08, the highly positive
1 8 0 ~.''q ......... i ........ i ........ - ........ i ........ i l 0 H/di = 7 flow exit angles (0 = +70 °, +55 °) achieve higher local
i ~ ~ ~ ~I ~ H/di = 3;[11
E 1 60 ~ : : - - - - . - i ........ i--I - - • - H/d! = 5;[1] Nusselt numbers (Fig. llb).
z il ° H/d,=7:tl] It is assumed that, in this respect, radial jet nozzles can
1 40 ........ i ......... ~... e~ ' .....: ........ i........ : ........ be treated like inline jet nozzles: optimal local heat trans-
fer is achieved when the dimensionless length L/b* of
i i
. . . . . . . . . : . . . . . . . .
the impinging free jet is equal to its potential core length
100..iii
.......i........ i
0 10 20 30 40 50 60
~ 70 80 90100
(L/b* = 5-8). Because of the curvature of the jet, L/b*
is always larger than the actual nozzle height, H/b*;
additionally, L/b* increases with decreasing flow exit
A / A N o = f-1
angle, 0. This explains that the optimal relative nozzle
Figure 9. Average Nusselt number, N u m , of inline jet flow height, H/b*, also decreases with decreasing values of 0.
plotted against dimensionless reciprocal nozzle area f-~ for The curves plotted in Fig. 12 show the course of the
different scaled nozzle heights H/di; Re = 54,000; curves: average Nusselt number, Num, over the reciprocal relative
correlation given by Schliinder and Gnielinski [1]. nozzle area for inline and radial jet nozzles, each posi-
200 F. Peper et al.

a) 400 The inline jet configurations with the highest local

Radiaijet; H / b . = 2.64 ( H i d i = 7) and highest average ( H i d i = 1) Nusselt num-
Re = 54000; vr,E = 30 rn/s
bers are a d d e d in Fig. 12. A comparison shows that, over a
300 ..............i...-.-~,,..,i .......................... O O = +70*
7-)~K ; ~ o = +s5* wide relative nozzle area 3 < f 1 < 70 the +55 ° radial jet
[ ~ [ a 0=+18.
• i A e = -50 nozzle achieves a b e t t e r heat transfer than does the inline
7= 200 ! ~ 0 O = -120 jet nozzle in optimal position. I m p r o v e m e n t of as much as
50% can be m a d e if a circular area of about ten times the
nozzle exit cross-section area is to be cooled, heated, or
1 O0 dried. Even radial jet nozzles with flow exit angles of
0 = +4 ° and 0 = +18 ° deliver an average heat transfer
i i i i that is c o m p a r a b l e to that of optimally placed ILJ nozzles.
00 4 8 12 16 20 24 However, radial jets with negative flow exit angles dry or
r/b ° cool less well than do inline jets.
The error in determining the local Nusselt number [Eq.
(11)] is equal to the error in determining the local heat
transfer coefficient, a , assuming that the errors measuring
b) 400 - Radial jet; H / b * = 4.08.
Re = 54000; VF,E = 3 0 m/s the inner nozzle diameter, di, and the thermal conductiv-
ity, AI;No, are negligible. The accuracy of calculating the
i _~i o 0=+700 local heat transfer coefficient, a [Eq. (9)], depends on the
300 ............... 1-3~/~----~ .............. ~......... --*- o = +55.
accuracy of measuring the t e m p e r a t u r e difference TFoH --
~/ "~.~ i • 0=-5* Tv,~. The uncertainty of measuring the specific convective
~- 200 heat flux, qrc....... is less than 1% and neglected. The
uncertainty of the local heat transfer coefficient, c~, has its
maximum at the impingement point because Tvoi~ - TF, E
1 00 has a minimum. Further, at the impingement point, a
systematic error of < 9% is m a d e by equating TF,~ with
i T~d,st [see Eq. (13)], which means that the presented local
00 4 8 12 16 20 24 values at the impingement point are as much as 9% lower
r/b* than the actual values. The total uncertainty at the stagna-
tion point was calculated to be less than 16%. The uncer-
Figure 11. Local Nusselt number, Nu, of radial jet flow
tainty of measuring the average heat transfer coefficient,
p lot t ed against dimensionless radius r / b * for different flow
a m, decreases with increasing size of the averaging area
exit angles 0; Re = 54,000. (a) H / b * = 2.04; (b) H / b * =
4.08.
from A a m _< + 1 6 % ( f -1 = 0, impingement point for in-
line jets) to Aar~ < + 3 . 1 % ( f 1 = 100).

WALL PRESSURE DISTRIBUTION

tioned at its respective optimal height. F r o m a comparison
of the radial jet nozzles in Fig. 12, it becomes clear that To present the results of the wall pressure distribution
the flow exit angle 0 = +55 ° provides the highest average measurements, a dimensionless pressure coefficient, Cp, is
heat transfer. The nozzle with 0 = +70 ° performs less defined:
well owing to its smaller r e a t t a c h m e n t radius, which re- Pstat,w -- P~
sults in a smaller area of high heat transfer. Cp- 0'5OF. E L ,2F, l - • (19)
The wall pressure distribution on the impingement surface
is shown in Fig. 13 for an inline jet and in Fig. 14 for a
+4°-radial jet. Figure 13 shows that inline jets generate a
• ILJ;H/d| = 1
250 / • ILJ;H/dl repulsing force on the impingement surface. The wall

~i
[1. e=+4.
I
X ® = +70o;H/b* = 4.08
i :: O O = +55=; H/b* = 4.0B
1.6
200 Ir i ......T"
i:........
": -: T""
-i : ~>
* ~--®0== --1
+12o:
8o:Hi H/b*
H/hi
~ 2 * i ==
2"04
2.042.04 1.4 _ _ - ~......... ~.............. In-line Jet .....
1.2 ~ i .............. R e = 5 4 0 0 0 ; v r . . = 3 0 m / . . ....

{.~ 0,8 ~ . ~ 0i "........................

1 1............. ............... H
!'"'"t l = l- ......................
.......................
/ d"-e-- i .....
,oo i .........i........ 0.6
............... - - - ' ~ * ' ~ - ~
~L,~x%
............... i-.---.] - - B - -
; / ~
H/d i = 3
HLd i= 5
I .....
i ~ Comparison: radial jet -- inline jet !
50 [~ .Re --54000; W,E = 30 m / s
0.4 ....................... - . . . . . . . . . . . . . . i t -e- 7 I .....

0 10
; • i

20
.

30
; , i

40 50 60
. ; . i . ;
70
, i
80
,
90'100
02 ZZ
0.0
A/ANo = f-1 1 2 3 4
r/ri
Figure 12. Average Nusselt number, NUm, of inline and
radial jet flow plotted against dimensionless reciprocal nozzle Figure 13. Pressure coefficient, Cp, of inline jet flow plotted
area f-~ for different scaled nozzle heights and flow exit against dimensionless radius r / q for different scaled nozzle
angles 0; Re = 54,000. heights H/d~; Re = 54,000.
 Impinging Radial and Inline Jets 201

pressure has its maximum value at the jet axis and de- C o m p a r i s o n : radial j e t - inline jet
1.5
creases with increasing radial distance in a bell-shaped
1.3 ...........~-.............. i..... O e--120
curve. The maximum values increase with decreasing noz-
zle height, H i d i. Owing to the spreading of the jet, the 1.1
0.9
...............F \
:\ i
........ I ............... I /
............... !------~ll.---J"~ ............ i----/
i / ,I,
r,
x
0---5o
e=
e = +280
+,,.
curve becomes somewhat broader with increasing nozzle .............. i .......~ / . : . \ ~ . i / o e=+ss*
to surface distance. The wall pressure distribution on the 0 . 7 .............. T"....... " ~ : V ' ~ ' " T / }K 0 = +700
.............. • ,,.,..,,°,=,
impingement surface by radial jets is different from that
0.3 ...............'.. . . . . . . . . . . . . . . i...............;..............
by inline jets, as shown in Fig. 14. The +4°-radial jet
generates a relative constant negative pressure in the area
beneath the flow guiding plate, r / b * < 4.08. With increas- -0.1
ing radial distance from the nozzle's centerline, the wall -0.3
pressure becomes more negative before there is a sharp - 0 " 5 0 ~ 6
rise from the negative to the positive peak, which coin- r/rl
cides with the position of the reattachment ring. For
larger values of r / b * , the course of the pressure coeffi- Figure 15. Pressure coefficient,
• .
C_,
1J
of . inline and radial jet
cients is similar to that of inline jets and decreases to zero flow plotted against dimensionless radms r / r i for different
in a bell-shaped curve. flow exit angles. ILl: H / d i = 1; RJ: H / b * = 4.08; Re =
The minima and the maxima are more distinct for 54,000.
small-scaled nozzle heights H / b * . With increasing nozzle
height, the reattachment radius increases; hence the jet The pressure coefficient [Eq. (18)] is a function of static
m o m e n t u m per unit perimeter and consequently the Cp pressure difference, A p s t a t , density, PF, E, and average flow
value at the reattachment ring decreases. Moreover, the exit velocity, UF.E. The errors measurmg A p s t a t and PF.E
entrainment effect becomes weaker if the volume beneath are less than 1% and have been neglected. The accuracy
the flow guiding plate is increased, reducing negative of determining the average flow exit velocity depends on
pressure coefficients within the reattachment ring. the accuracy of measuring the volumetric flow rate, l/F,
Figure 15 shows a comparason of wall pressure distribu- and the nozzle slot width, b. From this, the uncertainty of
tions for inline and radial jets at scaled nozzle heights of wall pressure distribution measurements was estimated
H / d i = 1 (ILJ) and H / b * = 4.08 (RJ). The pressure dis- according to the standard Gaussian method to be less
tributions of the radial jets emphasize that the reattach- than 7.9%.
ment radius increases with decreasing flow exit angle 0.
The minima as well as the maxima become less distinct Reattachment Radii In turbulent jet flow, the reattach-
with decreasing value of 0 for the same reasons men- ment radius is nearly independent of Reynolds number [4,
tioned earlier for variations of H / b * . Local negative wall 21]. Its location is determined by the maximum of local
pressure beneath the flow guiding plate is achieved for wall pressure, which coincides with the reattachment
flow exit angles 0 < +55 °. streamline.
From the pressure distributions presented in Fig. 15, it In Fig. 16, measured scaled reattachment radii, ( r R --
can be derived that the total jet force, F, exerted on the r o ) / H , are plotted against the flow exit angle, 0, for
impingement surface is higher for inline jets than for different average flow exit velocities, v v E. For positive
radial jets. Negative values of 0 can provide slightly nega- flow exit angles, the measured radii coincide well, whereas
tive jet forces owing to the growing subatmospheric pres- there is some scattering of the values for negative flow
sure region beneath the flow guiding plate. At 0 = +4 °, exit angles. This is probably due to the fact that the
the resulting jet force is approximately zero, which can be reattachment of the radial jet flow is less distinct at
desirable if stress-sensitive products are to be cooled, negative flow exit angles. The maximum of local wall
heated, or dried. pressure, which determines rR, spreads, thus increasing
uncertainty of the measured r R.

0.8 |

"l i
¢tli ~ ~
0.61- ............. ",........... -1,,~---Rodioldet; O = +40
'"1~ Re = 54000; VF,E = 30 m / s
....
-r
'

3
-

..~.
....i..! ~ [R e a t t a c h r n e n t Radii

0 ,,.,=30,,/,o,,,~
i
o-,t ............. t ......... , .... • ~l:~i - i / [] VF.IE-- 4 5 r n / s -- 1 . 5 W.No
i i-~. Q i I 0 vF.E = 6 0 m / l = 2 VF,Ne . .

,3 : ~
. . . . .

............. L ~ (r.-ro)/H.;, 1.804 e(-e/4o)

I
o.o

-0.2 "~.............. ~'" 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0 ~
-30 -20 -10 10 20 30 40 50 BO 70 80
0 4 8 12 16 20 24 O [o]
r/b"
Figure 16. Scaled reattachment radii of radial jet flow (r, -
Figure 14. Pressure coefficient, Cp, of radial jet flow plotted r o ) / H plotted against flow exit angle 0 for average flow exit
against dimensionless radius r / b * for 0 = 44 ° and different velocities VF, E = 30 m / s , 45 re~s, and 60 m/s; Re = 54,000
scaled nozzle heights H / b * ; Re = 54,000. = constant.
202 F. Peper et al.

However, an exponential correlation is suggested here the correlation of Blasius [22] for smooth circular ducts:
to estimate the reattachment radius from known values of Az 0.316
0, H, and ro:
,.oo = d-7 4R e- • (2s)
rR -- r° -- 1.804e o/4o. (20)
H The difference between fftheo and the measured values of
~'ILJ is < 4% in the considered Reynolds number range
Equation (20) is added in Fig. 16 for comparison. The 3 x 104 < Re < 6.104.
range of validity is
The RJ nozzle resistance coefficients, ~'RJ, serve to
- 2 0 ° < 0 < +40°: 2 < H / b * < 12 calculate the necessary additional fan power, Pr~, in the
+40 °< 0< +70°:2< H/b* <8 installation of radial jet nozzle systems:
• 1 2
PRJ = (26) ~ MF2UF, No"
P R E S S U R E LOSS M E A S U R E M E N T S By adding PRJ to a given dissipation of a technical appa-
ratus, the total dissipation or minimum required fan power,
Analysis Ptot, can be calculated according to Eq. (21).
For radial jet nozzles, the deflection of the jet at the flow
Results
guiding plate causes additional friction and possible flow
separation. Consequently, additional fan power is required In Fig. 17, the measured additional resistance coefficients
to maintain the same mass flow rate as that for an inline ~RJ are plotted against the flow exit angle 0 for the
jet. A dimensionless resistance coefficient, ~', is introduced precondition VF,E = rE, No [Eq. (2)], which was used
to evaluate this effect. Values of such resistance coeffi- throughout the comparison between inline and radial jet
cients are listed for many technical deviced such as valves, nozzles• Figure 17 shows that the resistance coefficient
orifices, flow straighteners, and so forth. The minimum ffRJ generally decreases with increasing flow exit angle.
fan power, /'tot, needed to overcome the total dissipation However, the values measured for flow exit angles 0 < +4 °
of the flow through a technical apparatus such as a paper somewhat differ from this rule, which is probably caused
drying machine can be calculated by adding the dissipa- by the different shapes of the nozzles and flow guiding
tions of all parts: plates (see Fig. 4) and hence different effects on the flow
pattern.
Ptot = EPi = ~ f f i M' 1i 2 c i 2 . (21)
The +55°-radial jet causes an additional resistance of
To evaluate the resistance coefficients ffRJ of the radial ffRJ ----"0.15. According to Eq. (26), this is equivalent to an
jet nozzle designs considered in this paper, a static pres- additional fan power of PRJ = 0.74 W per nozzle (assum-
sure port was installed in the nozzle wall at a distance of ing d i = 0.02 m; OF,No = 30 m/s). For many applications,
200 mm ("1") from the nozzle exit. The static pressure this value seems to be negligible if improvements in aver-
drop against ambient pressure, Apstat, l=, was measured. age heat or mass transfer up to 50% are achieved. Im-
The specific dissipation, q~l=, between the pressure port proved aerodynamical design of the flow guiding plates
and ambient air is might reduce the resistance coefficient further.
The resistance coefficient [Eq. (23)] is a function of
~U2 Aptot, 1~ static pressure drop, Apstat, l~ , density, PF, E, and average
= (22)
~DI~ 2 F,1 PF, E velocity in the nozzle, C'V,No. The error of measuring PF, E
is less than 1% and was neglected. The uncertainty of
At the nozzle exit, the static pressure is assumed to be measuring the static pressure drop is considered to be due
equal to the ambient static pressure; Pstat, E = Pstat,~" Fur- to the low absolute values, especially for RJ nozzles with
thermore, the kinetic energy of the free jet dissipates 0 > +55 °. From this, the uncertainty of measurements of
completely [22], With v F, 1 = VV,No, Eq. (22) can be written resistance coefficients was calculated according to the
as standard Gaussian method to be less than 8.4%.
A Pstat, 1~
(23)
~'=1+ PF, E 2 2.0 ! i ~ 1 .: ~ ~ ~ ~ !
2 UF, No
1.a I ........
~i~.'
........
::".............
t RJ: : :
resistance : :
coefficients : '

Here the constant value ~" = 1 is equal to the resistance '"1 ....... ........ t ---
coefficient of an ideal jet spreading from a nozzle into the ........ t ........ i ......... i ......... ......... ........ t ........ ..........
surroundings. ~.2 I-d"~ ....... i . t ........ i:-----~ ......... i ........ ~ ........ ".."........ i ..........
1 0 L nen CO~lCaq c~nlca=. ~ i i !
The resistance coefficient ~" defined by Eq. (23) consists -,,. -/ -~~i~T ........ [ ~ ~ l i i i ......... i ........ T ........ i ........ i .........
o.~ ........ ~ . . . . ~ . . - - , ......... , ......... i ........ ~. ........ , ..........
of the resistance of the inline jet nozzle, ~'~i~, which is
actually a circular duct, and the resistance of the different ........ ! ........ ......... i ........ i ........ - ........ i .........
flow guiding plates, ~'R~, necessary to deflect the flow into 0 4 - . . : : ~ ~ i

the radial jet: 0.2

o.o ~ ......... ~................
i ~......... i ......... i ......... ~ .... .~........ i ..........
~" = ~ u + ~'~a. (24) --20 - 1 0 0 10 20 ,30 40 50 60 70 80
® [o]
Because ¢ and ElLa were measured separately, it is possi-
ble to present resistance coefficients CRJ for different Figure 17. Additional resistance coefficients ~'RJ of radial jet
radial jet nozzle designs. The procedure was verified by nozzle flow plotted against flow exit angle 0 for vv, z = rE, No;
comparing the inline jet nozzle resistance coefficients with 3 × 104 < Re < 6. 104.
 Impinging Radial and Inline Jets 203

PRACTICAL SIGNIFICANCE/USEFULNESS 5. The required additional fan power is low (resistance

coefficient ~'RJ < 0.25) for RJ nozzles with flow exit
The results presented in this paper emphasize that radial angles of +45 ° < 0 < + 60 °, which give the best heat
jet nozzles bear a high potential of h e a t / m a s s transfer transfer performance. For most applications, the rise of
augmentation for many applications. Up to 60% higher energy costs is negligible as compared with the increase
values of the local and up to 50% higher values of the of h e a t / m a s s transfer.
average heat transfer coefficient are possible compared
with a standard circular inline jet of the same volumetric
flow rate and of the same average flow exit velocity. The
rise in regular costs for applying RJ nozzles with optimal NOMENCLATURE
flow exit angles of 0 = 450-60 ° seem to be negligible
A area, m 2
compared with the benefits for h e a t / m a s s transfer perfor-
mance. The costs for converting existing jet drying or b* effective nozzle exit width, m
cooling systems into the radial jet technique are low. An b actual nozzle exit width (= b*/cosO), m
inexpensive solution might be to employ tin cones as flow Cp pressure coefficient, dimensionless
guiding plates. Cp specific heat of air, J / ( k g K)
The total jet force on the impingement surface is lower E voltage, V
for radial jets than for inline jets. This is favorable if f relative nozzle area (= r~/r 2)
stress-sensitive products are to be cooled, heated, or dried.
g gravitational acceleration, m / s 2
Gr Grashof number [ = gl3(TFoil -- T~)/(t~v2)]
SUMMARY H nozzle to surface spacing (nozzle height), m
I current, A
The local heat transfer and wall pressure distribution
produced by single impinging radial jets on a plane sur- L length of free jet, m
face as well as the pressure drop in the nozzles were l characteristic length for calculation of free
measured for nine different flow exit angles ( - 1 8 ° < 0 _< convection, m
+70 °) and different nozzle heights. Comparative measure- { = sls2/[2(s I + s2)]},
ments with single circular inline jet nozzles were per- A;/ mass flow rate, k g / s
formed. The major results are: Ma Mach number [= VV, E/(KRFTv, E) °'5]
1. Radial jet nozzles with flow exit angles of +45 ° < 0 < n number of circular rings for averaging
+60 ° achieve optimal average heat transfer. At 0 = Nu Nusselt number [see Eq. (11) for definition]
+55 °, average heat transfer over a relative nozzle area P fan power, W
3 < f l < 70 is as much as 50% higher compared with p pressure, N / m 2
an inline jet. For RJ nozzles with positive flow exit Pr Prandtl number [= r/Cp/A]
angles, local heat transfer at the reattachment ring is as ~) heat flux, W
much as 60% higher compared with local heat transfer
,) specific heat flux, W / m 2
at the stagnation point of ILJ nozzles. Both nozzle
types are compared in their respective optimal geomet- R specific gas constant, m2/(s 2 K)
ric configuration, with the same volumetric flow rate, Re Reynolds number [see Eq. (5) for definition]
the same average flow exit velocity, and the same r radial distance from nozzle centerline, m
Reynolds number, Re = 54,000. ri inner nozzle radius, m
2. Radial jets produce optimal local and average heat ro outer nozzle radius, m
transfer if the dimensionless length of the radial free rR radius of reattachment ring, m
jet is L / b * = 5-8. This implies that the optimal di-
sl, s 2 dimensions of inconel foil, m
mensionless nozzle height H / b * increases with increas-
ing flow exit angle 0, owing to the more direct impinge- Svoil wall thickness of inconel foil, m
ment of the radial free jet. T temperature, K
3. The total jet force exerted on the impingement surface Tad adiabatic wall temperature, K
is higher for inline jets than for radial jets and, for Tv, Or temperature of air flow at orifice, K
radial jets, decreases with decreasing flow exit angle 0. V volumetric flow rate, m3/s
For negative values of 0, the jet force may be slightly c average velocity, m / s
negative (suction force) owing to the large subatmo-
spheric pressure region beneath the flow guiding plate. Greek Symbols
At 0 = 0°, a total jet force of approximately zero was c~ heat transfer coefficient, W / ( m 2 K)
measured, which produces a very gentle h e a t / m a s s A relative error
transfer performance. Hence compared with inline jets, e emissivity coefficient
much higher flow rates improving h e a t / m a s s transfer
~"resistance coefficient
can be applied without damaging a fragile surface.
4. Radial jets with strongly negative flow exit angles (0 < 77 dynamic viscosity, N s / m 2
- 2 0 °) reattach only for nozzle heights that are too 0 angle of flow exit (RJ nozzle) relative to plane
small for most practical applications ( H / b * < 2). surface, degrees
Hence 0 = - 2 0 ° marks the lower limit of a useful 0pl angle of flow guiding plate, degrees
employment of RJ nozzles. OR angle of reattachment, degrees
204 F. Peper et al.

K isentropic e x p o n e n t 8. Page, R. H., Owtowari, C., and Seyed-Yagoobi, J., Infrared

~Foil t h e r m a l conductivity o f i n c o n e l foil, W / ( m K) Images of Jet Impingement. Proc. 20th Int. Congr. on High Speed
v k i n e m a t i c viscosity, m 2 / s Photography and Photonics, SPIE, Vol. 1801, 1992.
9. Page, R. H., Axisymmetric Gas Jets: Surface Impingement Phe-
p density, k g / m 2
nomena. Proc. 14th Canadian Congr. of Applied Mechanics CAN-
Subscripts CAM, Vol. 1, pp. 10-19, 1993.
ad adiabatic 10. Thiele, E. W., Dautel, S. W., Page, R. H., and Seyed-Yagoobi, J.,
cond conduction A New Paper Machine Drying Concept: A Multifunctional, Ra-
dial-Jet Reattachment Nozzle Blow Box. TAPPIJ. 76(7), 203-207,
E nozzle exit
1993.
F air flow 11. Thiele, E. W., Seyed-Yagoobi, J., Page, R. H., and Castillo-Garcia,
fconv forced convection H., Enhancement of Drying Rate, Moisture Profiling and Sheet
frconv free convection Stability on an Existing Paper Machine with R JR Blow Boxes.
i index for s u m m a t i o n Proc. TAPP1 Paper Makers Conference, Chicago, pp. 223-228,
1995.
ILJ inline jet
12. Laschefski, H. Holl, A., Grosse-Gorgemann, A., Mitra, N. K., and
I l o w e r side of i m p i n g e m e n t surface
Page, R. H., Flow Structure and Heat Transfer of Radial and
m averaged Axial Jet Reattachment on a Flat Plate. ASME WAM, HTE 210,
No within the nozzle 123-131, 1992.
rad radiation 13. Laschefski, H., Cziesla, T., and Mitra, N. K., Influence of Exit
RJ radial j e t Angle on Radial Jet Reattachment and Heat Transfer. J. Ther-
mophys. 9(1), 169-174, 1995.
St stagnation point
14. Andrews, M., Modeling of Turbulent Radial Jet Reattachment.
stat static Int. Syrup. on Heat and Mass Transfer in Energy Systems and
tot total Environmental Effects, Cancun, Mexico, 1993.
u u p p e r side of i m p i n g e m e n t surface 15. Serafin, M., and Peper, F., Konstruktive Gestaltung und Test des
w wall Diisenversuchsstandes. Konstruktiver Entwurf 95-07, LWS,
atmospheric Ruhr-Universit~it Bochum, 1995.
16. Moffat, R. J., Describing the Uncertainties in Experimental
Results. Exp. Thermal Fluid Sci. 1, 3-17, 1988.
REFERENCES 17. Baehr, H. D., and Stephan, K., Wiirme- und StoffiJbertragung.
1. Schliinder, E. U., and Gnielinski, V., W~irme- und Stoffiibergang Springer-Verlag, Berlin, Heidelberg, New York, 1994.
zwischen Gut und aufprallendem Diisenstrahl. Chemic Ing. Tech- 18. Stewartson, K., On the Free Convection from a Horizontal Plate.
nol. 39, 578-584, 1967. Z. Angew. Math. Phys. 9, 276-281, 1958.
2. Schliinder, E. U., Kr6tzsch, P., and Hennecke, Fr.-W., 19. Huber, A. M., and Viskanta, R., Convective Heat Transfer to a
Gesetzm~issigkeiten der W~irme- und Stofftibertragung bei der Confined Impinging Array of Air Jets with Spent Air Exits. J.
Prallstr6mung aus Rund- und Schlitzdiisen. Chemic Ing. Technol. Heat Transfer 116, 570-576, 1994.
42, 333-338, 1970. 20. Lytle, D., and Webb, B. W., Secondary Heat Transfer Maxima
3. Viskanta, R., Heat Transfer to Impinging Isothermal Gas and for Air Jet Impingement at Low Nozzle-To-Plate Spacing. In
Flame Jets. Exp. Thermal Fluid Sci. 6, 111 134, 1993. Experimental Heat Transfer, Fluid Mechanics and Thermodynamics
4. Page, R. H., Hadden, L. L., and Ostowari, C., Theory for Radial 1991, pp. 776-783, Elsevier, New York, 1991.
Jet Reattachment Flow. AL4A J. 27, 1500-1505, 1989. 21. Sawyer, R. A., The Flow Due to a Two-Dimensional Jet Issuing
5. Martin, H., Heat and Mass Transfer Between Impinging Jets and Parallel to a Flat Plate. J. Fluid Mech. 9, 543-561, 1960.
Solid Surfaces. Adv. Heat Transfer 13, 1-60, 1977. 22. Gersten, K. Einffthrungin die Str6mungsmechanik. 5th ed., Vieweg
6. Martin, H., Impinging Jets. In Single-Phase Convective Heat Verlag, Braunschweig/Wiesbaden, 1989.
Transfer, Chap. 2.5.6, pp. 1-10, Hemisphere Publishing Corpora-
tion, Washington, 1983.
7. Page, R. H., Carbone, J., and Ostowari, C., Radial Jet Reattach-
ment Force. Exp. Fluids 8(5), 13-17, 1990. Received March 10, 1996; revised May 10, 1996