F. J.

BAYLEY The Fluid Dynamics of a Shrouded Disk System

Professor of Mechanical Engineering.

J. M. OWEN
With a Radial Outflow of Coolant
Rolls-Royce Research Fellow.
This paper describes an experimental study of an air-cooled gas turbine disk using the
Mechanical Engineering L a b o r a t o r y , model of a disk rotating near a shrouded stator. Measurements of pressure distribu-
School of A p p l i e d Sciences, tion, frictional moment, and the cooling air flow necessary to prevent the ingress of hot
University of Sussex, gases over the turbine disk are described for a range of rotational speeds, mass flow rates,
Sussex, England
and different geometries. The pressure distribution is shown to be calculable by the
superposition of the pressure drop due to the shroud and the unshrouded distribution.
Moment coefficients are shown to increase with increasing mass flow rate and decreasing
shroud clearance, but are little affected by the rotor/stator gap. Applying Reynolds
analogy to the moment coefficients, it is estimated that heat transfer from the rotor will
be controlled primarily by rate of radial cooling flow at low rotational Reynolds numbers,
and will be governed primarily by Reynolds number at large rotational speeds.

Introduction tion influences the thrust on the turbine rotor, while the drag
moment determines the windage loss and the dissipation energy
L
IT/THOUGH the combustion system and turbine in the air flow between rotor and stator, which may sometimes be
blades have posed the most critical cooling problems in the con- critical, particularly in high-pressure compressors. Both these
tinuing development of the gas turbine, the turbine rotor disk, parameters were measured experimentally and found to be in fair
also, has always needed some protection from the hot gases which agreement with theoretical predictions obtained from a numerical
pass axially over its periphery. Up to the present time com- integration of the parabolic boundary layer equations which were
paratively rudimentary techniques of disk cooling have been ade- used to represent this comparatively simple system.
quate, with the designer insuring only t h a t a radially outward In the present paper the earlier work is extended to the more
flow of air was maintained between the rotor and stator, using, as realistic situation, from the viewpoint of the practical turbine
necessary, peripheral glands or shrouds to insure no ingress of hot designer, in which a peripheral shroud is added to the system, as
gas at the perimeter. The continuing rise of gas temperatures in shown in Fig. 1(6). In a continuing program of research in the
the internal combustion turbine, however, with its consequent Mechanical Engineering Laboratories of the University of Sussex,
heavy demands on cooling air elsewhere in the cycle, has shown measurements have been made of the drag on the rotor, the radial
the inadequacy of our knowledge for predicting exactly the cool- pressure distribution, and of the detailed nature of the air flow
ing requirements of the turbine disk, both in determining heat within the system. A comparison has been made with the results
transfer rates over the faces of rotor and stator and, moreover, in from the earlier work with the unshrouded disks, and the probable
our fundamental understanding of the fluid dynamics of the effects on the coolant-disk heat transfer process are discussed.
flow.
The disk-cooling system in typical turbine practice is complex
but may be reduced in its simplest conception to the arrangement
shown in Fig. 1(a), in which a supply of cooling air from near the
axis of the system flows radially outward between the stationary
disk representing the nozzle diaphragm of the turbine and the
rotor disk. The fluid dynamics of this system have been studied
in detail by the present authors, and the effects of the radial out-
flow upon both the pressure distribution and the frictional mo- coolant
ment coefficient have been reported [1] .1 The pressure distribu- W

1
Numbers in brackets designate References at end of paper. stator.
Contributed by the Gas Turbine Division and presented at the
Gas Turbine Conference and Products Show, Brussels, Belgium,
May 24-28, 1970, of T H E AMERICAN SOCIETY OP MECHANICAL ENGI-
NEEBS. Manuscript received at ASME Headquarters, December 8,
1969. Paper No. 70-GT-6. Fig. 1(a) Unshrouded rotating disk model

Journal of Engineering for Power Copyright © 1970 by ASME JULY 1 9 70 / 335

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 shroud =1 f--sc

Fig. I(b) Shrouded rotating disk model

Fig. 2(0) Plate 4.1. General view of fluid dynamics apparatus

Experimental Apparatus
The apparatus used for the present work corresponds to the
diagrammatic representation of Figs. 1. The rotor is 30 in. in dia
and may be run at speeds up to 4500 rev/min at distances in-
finitely variable up to 3 in. from the coaxial stationary disk of the
same diameter. The rotor is made from mild steel and tapers
from a thickness of 11/4in. at the center to '/4 in. at the periphery
(approximately a 2-deg taper). For the experiments reported
here the stator was moulded from a '/.-in-thick transparent
acrylic sheet to the same shape as the steel rotor, and the radial
outflow of air was supplied at rates up to 2 lb/see from a com-
pressor through the 4-in. pipe opening into, and coaxial with, the
center of the stator.
The general arrangement of the apparatus is shown in the
photograph, Fig. 2(a) (Plate 4.1) while the details of the shroud
may be seen in Fig. 2(b) (Plate 4.2). This was constructed from
eight separate segments of acrylic sheet, in which were located
four axial static pressure tappings on the disk-side face, and four
rows of three radial tappings at '/s-in. intervals from the face.
Although the axial tappings consistently showed subatmospheric
pressure, because of the high local velocity in the gap between
Fig. 2(b) Plate 4.2. Perspox stotor with peripheral shroud assembled
rotor and shroud, it was found that the radial tapping gave a pre-
cise indication of the existence of outflow or inflow between the
surroundings and system. A subatmospheric pressure reading
from the radial tapping nearest the rotor showed that air was
entering the system and the superimposed radial flow was in- given mdius, within a few percent of each other, the flow was con-
adequate. sidered to be axisymmetric.
Variation in the clearance between the shroud face and the The drag moment on the rotating disk, which was driven by a
rotor was obtained by moving the shroud axially on the stator 15-hp d-c electric motor, was measured by a sensitive optical
disk to which it was attached. The stator also contained static torque meter built into the drive shaft. Tvleasurements of tlw
tappings on two orthogonal diameters, so that the distribution of drag torque and pressures in the system were taken for a range of
pressure in the shrouded system could be determined. As the imposed radial rates of air flow and rotational speeds, with the
readings on the four orthogonal lines of pressure taps were, for a clearance between shroud and rotor, 8" set at 0.05 and 0.10 ill.,

---Nomenclature----------------------------
A,B constant,9 in pressure rela- constants for mean radial He pWI'02/!J. rotational
tionship and tangential velocities R.eynolds number
lJi[o/('!.PW 21'0') = moment moment on enclosed side s = axial clearance between
coeft'icient of rotating disk stator and rotor
(p - PA)Pl'o2/jJ.2 pres- s,. axial clearance between
qo.avro/[le(To -- To""I)"v] =
sure coeft'icien t mean Nusselt number shroud and rotor
specific heat of fluid IV superimposed mass flow
W / jJ.ro = mass flow coef- P,PA static and atmospheric
rate
ficient pressure, respectively
To, T o,ad/'1', = disk, adiabatic disk, an,!
Cw,miu minimum mass flow to PI' jJ.elle = Prandtl number stator temperatures, r0-
prevent inflow mean heat flux through spectively
s/ro = gap ratio rotating disk jJ. fluid viscosity
s,1l'o = shroud clearance l' radial distance from disk p fluid densi ty
r~tio center line W angular velocity of ro-
k fluid thermal conductivity 1'0 outer radius of disk tating disk

336 / J U LY 1 970 Transactions of the AS ME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 experiment
<V 10+ G
c
experiment
VIQ4 °c
A 2.4 .0033 a 5.3 .006?
V 2.4 .0067 A 2.4 .0033

100C 2.4 no shroud V 2.4 .0067

o 0 .0033 2.3 no shroud

0 no shroud o 0 .0033
100 C, no shroud
mp 0

.6 .8 1. 6. 8.
•2 .4 .6 .8 1. ^ 2. 4. 6. 8. Re/10 6
Re/10 6
Fig. 3(b) Effect of shroud on moment coefficient for G = 0.12
Fig. 3(a) Effect of shroud on moment coefficient for G = 0.06

and the gap between rotor and stator, s, at

These values correspond, for a radius ro =
0.9, 1.8, and 2.7 in.
15 in., to clearance
experiment
V1"" K
A 2.4 .0033
ratios of Oc = 0.0033 and 0.0067, where Oc = sjr0, and 0 = 0.06,
0.12, and 0.18, where 0 = s/?'0. The results obtained from these V 2.4 .006?
experiments are discussed below.
2.4 no shroud

O 0 .0033
Drag Moment in the Shrouded System
Compared with the open stator-rotor system of Fig. 1(a), the 100 C, 0 no shroud
addition of a shroud will affect the drag on the rotating disk in the 'mp
absence of a superimposed radial outflow of cooling air in two
mutually contradictory ways. First, by restricting the flow area
at the periphery and accelerating the radial component of veloc-
ity, the shroud will increase the.viscid shear in the region where it
will have the greatest effect. On the other hand, by restricting
the ingress of air from the surroundings, angular momentum will
be conserved between the rotor and stator, so reducing the rela-
tive fluid motion and, hence, the drag on the rotor. I n Figs. 3
the dimensionless moment coefficient, Cmi0 = Af„/1/2pco2/'o6, is
plotted against the rotational Reynolds number, poir^Jy,. I t
can be seen t h a t the first of these effects predominates at low
Reynolds numbers where the entrainment of air into the rotating
boundary layer is small, while the converse applies a t higher
speeds. This effect is particularly evident in Fig. 3(a), where the
gap between rotor and stator was 0.9 in. (O = 0.06) and the
clearance, 0.05 in.. (Oc = 0.0033). As the gap was increased to
1.8 and 2.7 in., with the clearance unchanged, similar trends were
observed, as may be seen in Figs. 3(a), 3(6), and 3(c), out the
point a t which the drag of the shrouded system crossed the open .2 .4 .6 .8 1. 2. 4 6. 8.
system data moved to lower Reynolds numbers. With the larger Re/10*
gaps our earlier work on unshrouded systems had shown the in- Fig. 3(c) Effect of shroud on moment coefficient for G = 0.18
terior flow to be more quiescent, and thus more responsive to the
angular momentum conserved within the open system. Thus the flow at the periphery is now dominant, and the drag moment is
under these conditions the effect of shrouding, apart from the consistently increased compared with the open system, as shown
very lowest and unrealistic rotational speeds, is to reduce the drag in Figs. 3. Comparison of these figures shows also t h a t the
moment on the rotor. moment is very little affected by the gap between rotor and
The superposition of the radially outward coolant flow exactly stator, and the only significant geometric variable is the axial
reverses this situation, for.the effect of the shroud in accelerating clearance between shroud and rotor.

Journal of Engineering for Power JULY 197 0 / 337

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 experiment Re/10 6 k
A 0 no
experiment y#
shroud O 5.1.
O 3.6
A 0 A IX
.0067
0 3.6 ACp/10 D 1.6
A 0
.0033
cp/io ® 3.6
100 r

80 ~A A A A A A

-© ® & ®-
80
60

60
(.0
-A A A-
(.0
20
-@ © © ®~

.6 .7 .8 .9
4l—4 &—&—^g>—o 'IO1 •»=
r/r

-20 0 250 500 750 1000 1250

Fig. 4 Effect of shroud on pressure coefficient for G = 0 . 0 6 , C„ = 2.4 KeVlO"

X IO4
Fig. 6 Effect of Reynolds number on shroud exit pressure drop for G
0.18, Gc = 0 . 0 0 6 7

experiment 0
The Radial Pressure Distribution
O .06
Fig. 4 shows the distribution of static pressure, as measured on
A .1 2 the stator wall, represented as the pressure coefficient Cp = (p —
D .1 6 PA)P>V/H2- For these tests the gap between rotor and stator was
0.9 in., and in the absence of the shroud, with the rate of radial
outflow corresponding to the dimensionless flow coefficient C„ =
W/HTD having a value 2.4 X IO4 (W =• 0.4 lb/sec), the pressure
between the disks was slightly below atmospheric. A shroud
having a clearance of 0.10 in., Gc = 0.0067, as may be seen in Fig.
4, converts this to a large positive pressure even with the pump-
ing action associated with a rotational Reynolds number of 3.6
X IO6.
Close examination of the pressure distributions reveals that at
radii less than about r/?' 0 = 0.8 the gradients are not significantly
affected by shrouding, and a similar conclusion can be drawn
from the distributions at the other gaps tested. This suggests
t h a t if pressure distributions for the unshrouded situation are de-
termined, either experimentally or theoretically as earlier re-
ported by the present authors [1], then the effect of shrouding
on the absolute level of pressure may be determined by super-
position, for the "parabolic" regions of flow a t least, t h a t is, for
r/u less than 0.8.
The superimposed pressure coefficient due to the shroud,
AC„, say, may be determined from the present data, and examina-
tion of Fig. 4 shows it to be represented very closely by the
normalized pressure reading at r/r„ = 0.864. I n Fig. 5, AC,,
evaluated in this way is shown plotted against the square of the
mass flow coefficient when both disks were stationary; Fig. 6
shows the effect of rotation represented in the abscissa as the
C'/1Q 10 ' square of the Reynolds number. In each case the linear form of
w
the correlation is clear, as may be predicted theoretically.
Fig. 5 J Effect of mass flow on shroud exit pressure drop for Re = 0,G C Thus, if the radial pressure gradient is large compared with any
= 0.0067 axial variations near the periphery, and viscid forces are negligible

338 / JULY 1970 Transactions of the AS ME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 compared with the large acceleration forces, then the equation of Q . /1 rj
W min
motion becomes '

1 dp
= Vr ^ J : _ u * . (1)
p dr dr

It is further assumed t h a t the radial component of velocity at a ,•

radius r may be represented as a mean value given by

hW
Vr = (2)
2irprs 17 -

and at outlet from the shroud

V, - (2a) 1.0
2irpr0sc'

the tangential component is given by

V$ = kiO>r (3)
kt, ki, are constants.
The equation of motion can then be integrated between radius r
and the outer radius ?'o to give

AC',, - 'A
2TT / L^ 2 W G2J experiment G G
c
® .06 .0033
Vafa Re 2 !

HO- (4) 0
H
.06
.12
.0067
.0033
= ACw2 - BRe 2 (4a) 0 .12 .0067
A .16 .0033
The constants A and B contain the geometric parameters of equa-
A .16 .0067
tion (4), as well as the unknown constants, ki and kt, used to de-
fine the radial and tangential components of velocity, which also
might be expected to depend upon the geometry of the system. 1 2 3 t 5
Table 1 shows, however, that these parameters, derived from Re/10
the present experimental work for Gc = 0.0067, r/r0 = 0.864, do Fig. 7 Effect of shroud clearance on minimum mass flow
not vary very widely over the range of gap ratios tested and may
be used for design purposes with broadly similar geometries.

The difference between the constants in equations (5) and (6) lies
Table I
in the arbitrary definition of ACP in (4) and (5). As Fig. 4
G h h showed, the pressure at r/n = 0.864 sufficed to represent the in-
cremental pressure due to the shroud when the mass flow con-
0.06 0.66 0.21
siderably exceeded the minimum value to prevent ingress of
0.12 0.63 0.20
surrounding air. As this value is approached, however, either
0.18 0.58 0.17
by reducing Cw or increasing Re in equation (4), the role of the
shroud becomes more complex. The bulk of the outgoing fluid
Equation (4) indicates that at sufficiently high Reynolds num- migrates toward the rotor, so that the shroud presents less of a
bers, the pressure drop can go negative and this was certainly restriction to this, but begins to operate by impeding the inflow
observed in the present test program. I t does not follow, how- necessary to make up the rotating boundary layer entrainment.
ever, that a negative value of ACP, arbitrarily defined here as the I t is suggested, therefore, that equation (4) be used to estimate
normalized pressure difference between r/ro = 0.864 and r/r0 = 1, pressure distributions when the outflow is at least 30 percent in
represents the onset of inflow through the shroud clearance. A excess of the minimum radial flow required to prevent ingress,
separate study was made of the minimum rate of flow required to and that this flow be calculated from equation (5).
prevent ingress of air from the surroundings, C»,min, using the
criterion previously defined, that the pressure just under the disk-
side edge of the shroud should not become subatmospheric. The Heat Transfer in the Rotating Disk System
results are shown in Fig. 7, from which it appears that the shroud Dorfmau's [2] application of the Reynolds analogy for heat
clearance is the controlling geometric parameter, and t h a t the transfer from a free disk has been extended by Owen [3] to the
stator-rotor gap is not important in the range tested. Previous case of a disk rotating near a stator with a radial outflow of air
tests with unshrouded stators showed that C«,mm increased with (the model shown in Fig. 1(a)). For the case of unity, Prandtl
increasing gap ratio, and so it is apparent t h a t the gap ratio will number and a quadratic temperature rise over the rotating disk
become significant in shrouded systems when G and Gc are of (such that the disk temperature, T0, is proportional to r 2 ), the
comparable size. mean Nusselt number, Nu, is given by
The results in Fig. 7 are correlated approximately by
Nu Re Cm,Jv (7)
C„ 0.61(?oRe (S)
where
An exactly similar form is obtained by putting ACP in equation
(4) equal to zero, but with the mean values for ki and ki from g°,av r„
Nu = (8)
Table 1, there results

0.86(?cRe (6) cjo,a.v being the mean heat flux through the rotating disk, and A the

Journal of Engineering for Power JULY 1970 / 339

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Cw/104 Nu/10'
Author
7.6 5.2 2.3
6.
Kreith et al U m 01
mass transfer
Mitchell . |
—Q>-- .-©— —0>—
heat transfer
Kapinos (6 ---©-- -H&-- -®---
heat transfer
Bayley &Qwen ( 1 \ -¥-
moment coefficients

&c c w /io +
0.OOB7 5.3

lOO'C,
0.0033 2.1
mo
2. r 0.0067 2.1

_l I L_

.1 .6 l. 6.
Re/10*

Re/10 Fig. 9 Reynolds analogy applied to shrouded disk: 6 = 0.12

Fig. 8 Effect of mass flow rate on moment coefficients and Nusselt of a rotor-stator system with radial outflow, For air, Pr = 0.72,
numbers for G = 0.12: no shroud equation (7) can be written
Nu = 0.261 Re Cm,„ (10)
thermal conductivity of the fluid. Where Motional heating is
negligible, the adiabatic disk temperature, To,ad, is equal to the The moment coefficients for a gap ratio of O = 0.12, together
stator temperature, T,. For large rotational speeds, where dis- with the corresponding Nusselt numbers calculated from equation
sipative heating of the rotating disk can be significant, the (10), are shown in Fig. 8. While the Nusselt numbers are larger
adiabatic temperature is given by than those of Mitchell and Kapinos, the results do show a
transition between these two ranges. The Nusselt numbers ob-
r„,ad = T3 + V^VVC (9) tained from the correlation of Kreith, et al., for a Reynolds number
of 105 are not inconsistent with the trends shown from the
C being the specific heat of the fluid. For axial-flow gas tur-
Reynolds analogy.
bines, where Reynolds numbers up to 107 are common, the dy-
namic temperature in equation (9) can be as high as 100 F. For shrouded systems, the Reynolds analogy is not justifiable
in mathematical terms, owing to the fact that the boundary layer
For the case of radial outflow between a stator and a disk rotat-
assumptions are no longer valid. However, a qualitative esti-
ing at relatively low Reynolds numbers (4 X 104 < Re < 105), the
mate of Nu should be obtained by applying the analogy to the
mass transfer experiments of Kreith, Doughman and Kozlowski
measured moment coefficients for the shrouded system. Fig. 9
[4] showed that heat transfer depended primarily on the radial
shows the Nusselt numbers calculated from the moment co-
mass flow, and very little on rotation. Mitchell [5], operating at
efficient given in Fig. 3(6) for O = 0.12.
Re < 7.5 X 105, found that the Nusselt number was virtually in-
dependent of Re, but was strongly dependent on the radial mass It should be pointed out that, while the Nusselt number will be
flow. By contrast, Kapinos [6] produced results which showed increased for decreasing shroud clearance, so also will the Mo-
that for 5 X 106 < Re < 4 X 106 the Nusselt numbers were a tional heating effect. A program of heat transfer research is
strong function of Re, and were little affected by the mass rate of currently in operation at the University of Sussex, and it is in-
radial flow. Nusselt numbers calculated from a Reynolds tended that experiments on heat transfer from the unshrouded
analogy with the authors' own measured moment coefficients disk will be extended to Motional heating and heat transfer from
tend to reconcile the two extreme views: at large Re, Nu is de- the shrouded disk.
pendent primarily on Re; at low Re, Nu is dependent primarily
on(7„. Conclusions
The Reynolds analogy is based on equation (7) modified for This paper has described the experimental study of an air-
Prandtl number, Pr, by Dorfman's correction cooled gas turbine disk using the model of a disk rotating near a
shrouded stator.
Nu (Pr) = Pr»' 6 Nu (Pr = 1)
The principal conclusions to be drawn from this work are:
Although this correction was made for a free disk, it has been 1 For the case of no superimposed radial outflow, the moment
shown by Owen [7] to apply with reasonable accuracy to the case coefficient, compared with an open unshrouded system is, for

340 / JULY 1970 Transactions of the AS ME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Reynolds numbers above Re = 106, reduced by the presence of a for a disk rotating near a shrouded stator is expected to be in-
shroud through the conservation of angular momentum between creased by decreasing shroud clearances, but should be largely
rotor and stator. For lower Reynolds numbers, the shroud does independent of the gap between rotor and stator. While the
not impede the ingress and egress of entrained fluid, and the ac- Nusselt number will be increased by raising the radial mass flow
celeration through the relatively small shroud clearance causes rate, the improvement will be attenuated at increasing Reynolds
increased frictional moments. numbers. The increase of Nusselt number for the shrouded
2 For forced radial outflow, the presence of a shroud causes system compared with the unshrouded system will, to some ex-
increased moment coefficients for all Reynolds numbers, com- tent, be offset by the increase in dissipative heat generated by the
pared with the unshrouded system, owing to the increase of shear friction due to the shroud.
stress at the outer radii as the fluid passes through the shroud
clearance. Although the moment coefficient increases with de- Acknowledgment
creasing shroud clearance, it is—for the range of geometries
Our thanks are due to Rolls-Royce Ltd. for their continuing
tested—little affected by the gap between rotor and stator.
support of the work reported in this paper.
3 The radial distribution of static pressure is, for r/r„ < 0.86,
insensitive to the presence of a shroud, which, however, sig-
nificantly alters the absolute level of pressure within the system.
References
The pressure distribution can be regarded as the superposition of 1 Bayley, P. J., and Owen, J. M., "Flow Between a Rotating and
a Stationary Disc," Aeronautical Quarterly, Vol. 20, 1969, p. 333.
the unshrouded distribution, calculable from solution of the 2 Dorfman, L. A., "Hydrodynamic Resistance and Heat Loss of
boundary layer equations [1], and the pressure drop across the Rotating Solids," Oliver and Boyd, Edinburgh, 1963.
shroud, AC,,, where 3 Owen, J. M., "The Reynolds Analogy Applied to Flow Between
a Rotating and a Stationary Disc," School of Applied Sciences, Uni-
ACP = AC,. 2 - B Re 2 versity of Sussex, Mechanical Engineering Report 69/Me/16.
4 Kreith, F., Doughman, E., and Kozlowski, H., "Mass and Heat
Transfer From an Enclosed Rotating Disc With and Without Source
A and B being empirical constants. Flow," Journal of Heat Transfer, TRANS. ASME, Series C, Vol. 85,
4 The minimum mass flow necessary to prevent ingress of air No. 2, May 1963, p. 153.
from the surroundings, C«i,min, was found, for the range of 5 Mitchell, J. W., "A Study of the Fluid Dynamics and Heat
geometries tested, to be strongly dependent on the shroud Transfer Behavior for Radially Inward Flow Over a Shrouded Ro-
clearance, but very little affected by the rotor/stator clearance. tating Disc," TR No. 57, Department of Mechanical Engineering,
Stanford University, Stanford, California, 1963.
The following empirical relationship is proposed for the case of 6 Kapinos, V. M., "Heat Transfer From a Disc Rotating in a
Gc« G, Housing With a Radial Flow of Coolant," Journal of Engineering
Phyics., Vol. 8, 1965, p. 35.
C„,min = 0.61<?c Re 7 Owen, J. M., "The Effect of Forced Flow on Heat Transfer
From a Disc Rotating Near a Stator," School of Applied Sciences,
5 Based upon the Reynolds analogy, the Nusselt number University of Sussex, Mechanical Engineering Report 69/Me/17.

Journal of Engineering for Power JULY 1 9 70 / 341

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 01/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use