NATIONAL ADVISORY COMMITTEE

FOR AERONAUTICS

TECHNICAL NOTE

No. 1587

INVESTIGATION OF EFFECTIVENESS OF AIR-HEATING A HOLLOW

STEEL PROPELLER FOR PROTECTION AGAINST ICING

II - 50-PERCENT PARTITIONED BLADES

By Porter J. Perkins and Donald R. Mulholland

Flight Propulsion Research Laboratory Cleveland, Ohio

Washington May 1948

NATIONAL ADVISORY COMMITTEZ FOR AERONAUTICS

TECHNICAL NOTE NO. 1587

INVESTIGATION OF J0 71FEWIVE10SS OF AIR-BETING A HOLLOW

STEEL PROP1TTRR FOR PROTECTION AGAINST ICfl

II - 50-PERCENT PARTITIONED BLADES

By Porter J. Perkins and Donald R. Mulholland

The icing protection afforded 5fl\ interna]. air-heated propeller blade by radial partitioning at 50-percent chord to confine the heated air to the forWard half of the blade was determined in the NACA Cleveland icing research tunnel. A modified production-model hollow steel propeller, was used for the investigation. Temperatures of the blade surfaces for several heating rates were measured under various tunnel Icing' conditions. Photographic observations of ice formations on blade surfaces and blade heat-exchanger effectiveness -were obtained.

With 50-percent partitioning of the blades, adequate Icing pro-tection at 1050 rpm was obtained with a heating rate of 26,000 Btu per hour per blade at the blade shank using an air temperature of 4000 F with a flow rate of 280 pounds per hour per blade, which is one-third less heat than was found necessary for similar Ice pro-tection with unpartitioned blades. The chordvise distribution of the applied heat, as determined by surface temperature measurements, was considered unsatisfactory with much of the heat dissipated well back of the leading edge. Heat-exchanger effectiveness of approxi-, niately 56 percent also Indicated poor utilization of available heat. This effectiveness was, however, 9. percent greater than that obtained from unpartitioned blades.

INTRODUCTION

The protection of propeller blades from icing by the internal passage of heated air has been previously Investigated in flight and in the laboratory. These Investigations are reported in references 1 and 2 (unpartitioned blades) and the results of a theoretical analysis are given in reference 3 (50-percent radially partitioned blades).

NACA TN No. 1587

Unpartitioned blades were found by experiment to require large quanti-ties of heat for icing protection and afforded an uneconomical chord-vise distribution of the heat.

In an effort to develop propeller blades requiring less total heat for icing protection, an investigation was conducted in the NA.CA Cleveland icing research tunnel using a modified air-heated hollow steel propeller with blades partitioned radially at 50-percent chord to confine the flow of heated air to approximately the forward one - ha].f of the blade cavity.

Data were obtained to determine the effectiveness of the blade heating system, blade-surface temperatures, and heat requirements. Observations were made of the icing protection attained by the heat-ing system.

APPARATUS AND

INSTRUMENTATION

The Icing Investigation was conducted in a return-type tunnel in which icing conditions were simulated by means of water sprays and the use of refrigerated air. The propeller was driven by a liquid-cooled engine mounted in an airplane fuselage located in the dif-fuser of the tunnel (fig. 1). A modified production model four-, bladed hollow steel propeller 10 feet, 2 inches in diameter was used. Heated air was distributed to the propeller blades, as shown in fig-ure2.

The instrumentation provided for measurement of tunnel and pro-peller operating conditions, propeller heated-air temperature and mass flow, and blade-surface temperatures. Chordwise blade-surface tem-peratures were measured, at 40- and 70-percent blade radii on two of the four blades using iron-constantan ribbon-type thermocouples spot-welded to the blade metal. The locations of blade-surface thermo-couples at 40- and 70-percent radii are shown in figure 3. In addition, leading-edge surface temperatures were measured at 30- and 55-percent radii. Further details concerning apparatus and instru-mentation used for the propeller-icing research are given in ref-

erence 1.

Blade partitioning. - The partition installed within the blade cavity was of semiregid fiber-glass construction. It was anchored and sealed at the blade shank and was held in position radially along the blade by means of rivets through the blade walls. A semi-circular construction was used (fig. 3), which held the partition

NACA TN No. 1587 3

firmly against the inner walls of the cavity between rivets to pre-vent air leakage to the rear portion of the blade cavity. The partition terminated approximately 6 inches from the blade tip to allow the heated air to pass to the rear of the blade and escape through a tip orifice on the camber face.

Propeller-blade heating system. - For the 50-percent parti-tioned blades, the air flow through the blades and the arrangement of seals were unchanged from those described in reference 1 for the; unpartitioned blades, except that the air was confined to the forward half of the blade. The tip orifices used for the 50-percent partitioned blades were approximately two-thirds the cross-sectional area of those used for the unpartitioned. blades. Construction details and photographs of the blade-shank inlets and tip orifice are shown in figure 4.

Heated air was supplied to the propeller through electric heaters outside the tunnel. The supply of heated air was manually controlled and the blade airflow was calculated by adding or sub-tracting the leakage past the graphite seals, depending on the seal differential pressure, from the measured manifold air flow. The calibration curve used to 'determine leakage values Is given in reference 1.

CONDITIONS AND PROCEDURE

Tunnel conditions. - The propeller was subjected to icing con-ditions at tunnel-air temperatures between _40 and 230 F. The tunnel-air velocity averaged approximately 200 feet per second measured at 75 percent of the propeller radius. During the icing period, the velocity decreased because of pressure losses resulting from tunnel ice formations. Liquid-water concentrations were approximately 0.1 gram per cubic meter at 30 F and 0.9 gram per cubic meter at 190 F. The effective droplet diameter based on the volume maximum averaged approximately 55 microns for all tunnel conditions. This value is larger than average droplet diameters usually found in the atmosphere. The icing of propeller blades, however, is not greatly reduced unless relatively small droplet diameters are encountered, because the collection efficiency of small bodies Is approximately 100 percent. For the conditions used, the heating values required in this investigation are considered conservative.

Range of propeller variables. - The propeller was operated at two blade angles, 280 and 350, measured at 75-percent radius with corresponding rotational speeds of 1050 and 850 rpm, at which

4

NACA TN No. 1587

advance ratios of 1.1 and 1.4, respectively, were established to approximate the advance ratio of peak efficiency. Under tunnel operating conditions, the heated-air flow per blade was adjusted to range between 250 and 330 pounds per hour at 1050 rpm and between 165 and 185 pounds per hour at 850 rpm with an air tem-perature at the shank between 3000 and 4000 F.

Typical procedure. - Data were obtained during 10-minute periods of simulated icing conditions. Prior to each icing period, data were recorded for 2 minutes with all operating conditions stabilized and heated air flowing through the blades. Blade-surface temperatures 'were recorded at approximately 1-minute intervals throughout the preicing and icing period. Following each icing condition residual icing photographs were obtained.

Blade Heating System

The blade heat-exchanger effectiveness, as defined in ref-erence 1, is the ratio of the heat dissipated from the blade sur-face to the heat applied, including the heat from adiabatic com-pression of the internal air due to rotation. The values of heat-exchanger effectiveness shown in figure 5 for both the 50-percent partitioned and the unpartitioned blades are for the same range of air temperatures entering the shanks (3000 to 4000 F). With no sprays In the tunnel and at a propeller speed of 1050 rpm, the effectiveness for 50-percent partitioned blades averaged 56 percent or 9 percent higher than that for the unpartitioned blades (fig. 5). The partition guided the heated air through the forward section of the blade, where external heat-transfer coefficients are highest, and, in addition, a heat sink was provided behind the partition causing much conduction to take place rearward through the blade metal as indicated by high surface temperatures aft of the partition.

The effect on the blade heat-exchanger effectiveness of the primer coating used to hold the surface thermocouples to the blade is also indicated in figure 5. The effectiveness of the painted blade was 6 to 7 percent lower than that of the bare blade.

Blade-Surface Temperatures

The temperatures measured by the thermocouples located on the blade surfaces are plotted as a rise in temperature above the ambient tunnel temperature (figs. 6 to 10). Because values of heating rate

NACA TN No. 1587

5

surface temperature, and tunnel temperature varied during an icing period when surface temperature data were taken, all the distribution curves were constructed from plots of average values of surface tem-perature rise above tunnel ambient temperature. In developing these average curves; all the temperature data for a given location were plotted against heating rate. By assuming a linear relaticn between the two.variables, a straight line was drawn as an average of the, data, from these plots the faired distribution curves were con-structed for various heating rates. For the range of data investi-gated the relation of surface temperature rise to heating rate was considered to independent of tunnel-air temperature.

The distribution of this temperature rise is shown in figure 6, where the blade chordwise surface is laid out in plan form with the leading edge as the center and the surface distance over the thrust and camber faces extending along the axis on either side. Data are presented. for a propeller speed of 1050 rpm and heating rates of 18,000 and 22,000 Btu per hour per blade with no water spray in the tunnel and heating rates of 22,000 and 28,000 Btu per hour per blade under icing conditions.

It is significant that partitioning the blade at 50-percent chord failed to concentrate most of the heat at the leading edge where it was most needed for icing protection. The high temperature rises aft d the leading edge and near the partition, especially at the 40-percent radius, indicated the dissipation of large quantities of heat in areas not subject to severe ice formations.

Increasing the heating rate from 18,000 to 22,000 Btu per hour per blade with no sprays in the tunnel resulted in practically uni-form ohordwiaè surface-temperature increases at both the 40- and 70-percent radii (figs. 6(a) and 6(c)). Under Icing conditions, an inórease in the heating rate from 22,000 to 28,000 Btu per hour per blade considerably increased the surface temperatures at 40-percent radius except near the leading edge (fig. 6(b)) but did not appreciably increase the temperatures at the 70-percent radius. The increased, heating rate did., however, materially Increase the temperature gradient aft of the leading edge on the inboard blade sections as shown in figure 6(b). The steeper gradlent'narrowod the potential Icing area and thus afforded greater icing protection for the blade.

The temperature rises at the 40- and 70-percent radii under no-spray and icing conditions at a heating rate of 22,000 Btu per hour

-per blade and. a propeller speed of 1050 rpm are compared In fig-ure 7. With no sprays, temperatures at the 40-percent radius lead-ing edge were considerably higher than those at the 70-percent radius.

NAQA TN No. 1587

The effect of water impingment was to lower all blade-surface tem-peratures. Under icing conditions, the temperatures near the lead-ing edge for the two radii were approximately equal. Aside from the effects of throw-off due to centrifugal forces, it can there-fore be seen that the two radii will be about equally susceptible to icing at the leading edge.

Data taken at a propeller speed of 850 rpm with a heating rate of 14,000 Btu per hour per blade are shown in figure 8 for the chord-wise temperature distributions at the 40- and 70-percent radii under icing conditions. Although the heating rate had been reduced from 22 1 000 to 14,000 Btu per hour per blade and the propeller speed from 1050 to 850 rpm, a comparison of figures 7 and 8 shows that the leading-edge temperature rises at the 40-percent radius were approximately equal. The temperature gradients for the two condi-tions were also similar. At the 70-percent radius, however, the leading-edge temperatures were lower than the 40-percent radius temperatures and were also lower than the corresponding 70-percent radius temperatures on the high-speed blade with the higher heating rate. Aside from throw-off effects, these results show that at 850 rpm with 14,000 Btu per hour per blade, the 70-percent radius would ice over the entire chord at higher ambient temperatures than the 40-percent radius.

Surface temperatures obtained on the leading edge at the 30-1 40-, and 55-percent radii for a propeller speed of 1050 rpm are presented as temperature rise above tunnel ambinet temperature in figure 9. The decrease in surface temperature rise with increasing radius results from the drop in internal gas temperature with increasing radius and the increase in external heat-transfer coef-ficient with increasing peripheral velocity. The increase in kinetic heating with peripheral velocity tends to counteract the two effects. These considerations together with results given in reference 2 indicate that with a higher propeller speed the tem-perature rise for the icing condition may begin to increase beyond the 55-percent radius. This temperature rise increase is cor-borated by the data of figure 7 where the leading-edge temperatures at the 40- and 70-percent radii under icing conditions differ by only 20 F. The opposite result may be observed In figure 8 where at a lower propeller speed of 850 rpm the 70-percent radius leading-edge temperature is 4 0 F lower than the corresponding 40-percent radius temperature.

A comparison of the surface temperature distributions obtained with 50-percent partitioned blades at a heating rate of 28,000 Btu per hour per blade and temperature distributions obtained, with unpar - tioned blades at a heating rate of 40 2 000 Btu per hour per blade is shown In figure 10. These data were obtained at a propeller speed of 1050 rpm under icing conditions. Some improvement in heat

NPLCA TN No. 1587

7

distribution and economy is found with the 50-percent partitioned blade. Temperatures at the leading edge for the 40- and 70-percent radii on the 50-percent partitioned blade are both equal to the leading-edge temperature at 70-percent radius on the unpartitioned blade, although only three-fourths of the heating rate used in the unpartitioned blade was used in the partitioned blade.

The high surface temperatures and, hence, ice protection aft of the partition in the 50-percent partitioned blades indicate that the back surfaces of the blade would be adequately protected from ice if the partition were moved even farther forward. By such a partitioning, ice protection might be afforded the entire blade with the use of a lower heating rate.

Icing Observations

Comparative photographs of residual ice formations for different heating rates are shown in figure 11. The heating rate was increased from 20,000 to 26,000 Btu per hour per blade at a propeller speed of 1050 rpm and a tunnel ambient-air temperature of 110 F. Increas-ing the heating rate to 26,000 Btu per hour per blade, corresponding to approximately 280 pounds of air per hour at a shank temperature of 4000 F, did not completely prevent ice formation on the leading edge of all the blades, as would be expected. The temperature dis-tributions shown in figures 6(b) and 6(d) show little increase near the leading edge when the heating rate was increased from 22,000 to 28,000 Btu per hour per blade. Lesä ice was to be found on the blades with the higher heating rate, however, because of the greater throw-off made possible by the steeper temperature gradients pre-viously discussed. The residual ice formations shown for the higher heating rate are not considered sufficient to impair the aerodynamic performance of the propeller. It had been established in reference 1 that in the tunnel, icing conditions were most severe at an ambient-air temperature of 110 F, and inasmuch as the blades were adequately protected for this condition by a heating rate of 26,000 Btu per hour per blade, this value is believed to be adequate for all other icing conditions. This heating rate is approximately two-thirds that required for similar protection using unpartitioned blades.

SUMMARY OF RESULTS

From an investigation conducted in the icing research tunnel, the following results of the icing protection afforded a propeller blade by the internal passage of heated air through the forward half of the blade were obtained: -

8 NACA TN No. 1587

1. Adequate icing protection existed at 1050 rpm with a heating rate of 26,000 Btu per hour per blade at a shank air temperature of 4000 F, and a flow rate of 280 pounds per hour per blade, which was about two-thirds the heating rate required for an unpartitioned. blade. Small ice accretions at the leading edge on the inboard blade sections were not considered significant with respect to aerodynamic performance of the propeller.

2. The chord.wiee distribution of the heat, although better than that of the unpartitioned blade, was still unsatisfactory because too much of the heat was dissipated on the rear sections of the blade and too little on the leading edge.

3. The heat-exchanger effectiveness of the 50-percent partitioned blade averaged 56 percent, which is approximately 9 percent more than that for the unpartitioned. blade.

Flight Propulsion Research Laboratory, National Advisory Committee for Aeronautics,

Cleveland, Ohio, January 19, 1948.

4L$;fcJ 1.. Mulholland, Donald R., and Perkins, Porter J.: Investigation of

the Effectiveness of Air-Heating a Hollow Steel Propeller for Protection Against Icing. I - Unpartitionéd Blades. NPCA TN No. 1586, 1948.

2. Darsow, John F., and Selna, James: A Flight Investigation of the Thermal Performance of an Air-Heated. Propeller. NACA TN No. 1178, 1947.

3. Gray, V. H., and Campbell, R. G.: A Method for Estimating Heat Requirements for Ice Prevention on Gas-Heated. Propeller Blades. NACA TN No. 1494, 1947.

H

0

I

NACA TN No. 1587

Page intentionally left blank

jo

Page intentionally left blank

NACA TN No. 1587

1<I

V

I0

43

I

4

g

CIS

cM

a 0

U, —4

ci "U

ci

I0

.1.)

C,

0

U

0

0 4.3

C, 0

0

0 C, 0

0

4.)

V 0

0

0

12

NACA TN No. 1587

Iowa

0.30" rad'

0.75" radf

0,55"-"4

0.79"

Pip 0 0

_q^_

Total area, 0.58 aq in.

0,37"

NACA TN No 1587

'3

Fillet line

Camber face

Tip orifice

Blade markShank inlets

(a) Construction details.

Figure 4- Shank inlets and tip orifice for 50—percent partitioned air—heated propeller blades.

Page intentionally left blank

Page intentionally left blank

LI Ln

0

"4'

0

-4

a j

I I 0

•0

-4 a 0

4)

I

0

I

N Zf -

Uc'

NACA TN No. 1587

15

< N <2 N

0 u CY

Page intentionally left blank

Page intentionally left blank

20,000 240000 28,000 52,000 36,000 40,000 40

16 0000

60

44

o Bare blade

E3 Painted blade

-----

50-pErcent par itio ed blade

0

XE3

/ -

7Unpa blate

titined (bare)

NA CA

56

43

C)

4, 01

C

'-4 43

C)

4, C-

52

4E

NACA TN No. 1587

17

Heating rate, Btu/(hr)(blade)

Figure 5.- Comparison of heat-exchanger effectiveness for unpar-titioned and 50-percent-chord partitioned air-heated oropeller blades. No sprays; propeller speed, 1050 rpm; tunnel air temperature, _30 to 150 F.

OD

qv

U N

4)

a) C)

:;

tO

0

cd

4)

"-I

4)

a>

4)

a>

ft a> 4)

C,

4)

£

C) aiO Ii r4 U') 0 P0 0. .-4

•1 o o

wo 4

•

Q4

14 bO

0. 00 '3 q-4,-4

4) ,-1 o lz a,

'1)0.

—a> U)

C's C)

C,

4-4

0

4) C,

'4

111

4)

18

NACA TN No. 1587

4 4l

7

cd ad --d 0 C 1-----4----

I 14 F-I e. a I

'-4 I I -

-

cd

-

I -4) I

-- V- ---

0

—.-

4) V cd 0••,

__

-k-- —s------

0

—Cd

P8G

7z / —7

ley

Oe 00'

43

-- ---

0 0 0

UP

CC) to lw LV

d0 l esTj 9JflBJ9dWL

NACA TN No. 1587

19

_

-

0o I -

QS 00 1c

I

-

-

• ' C,

'4

' \ V

\ \ . --

0

---

-

-•

--

-

----- -- +-

-

--.-•--

_L

- __-v- --

d0 1 9911 arnqvjodmeL

C?

'.4.

• .4uj

o d 4)r-4

co

43 V 4) to Ø

1

0 43V Cd

ce

V o •

0. .0

) '4 U v-I i 0 'Ui-I

'J V

V t C?4 V

0 C,

'0 H •ri

o -44 0 ,O O.r4

o - 43

010 V .i•4

%4

'.4 '0

I

20

NACA TN No. 1587

MENNEN Ui'

MENJEENMEN,_FAD_

-Ui

MEMPOES UUMMANS

rd

MEMEMEMENNEN EMILVEME MEMENE NOMINEE

I.

_

EMEN MEMMI mmaimm • . ________ • S

:

•, m

HEM MEN

Emmiskm MENEM

rn MEMME

MENEMrn

Em'som_

OEMMN MENEM

.4 Id

4)

0 C) £4 C

Na

C

£4

C

0

C,

C

cD 43

£4 0

0

0 CO 0

r C Id

.00.

00 It)

00 O 4)r4 .1 £4a •d •d

43 43 4)0 C '4 • •4 Id o

£4 1-4 • 0 o '0r4

0 0. 00

44 0 d 0 N

C a 0 tiO

C '.4. C .4

0 440 03 H 0 4)

— on d O'4 .-. 14f4

4443

* Id

C

C,

0 0

0

CO

0 £4

•14

co L) (.) Li LI LI

1.0 03 10 103

jo 'agIJ a..xneodweJ

o-4 C)

I ollZ

--- -

-40 0 0 -

I I,

---------\-

'5'..

'5'V C.)

N

—e--

IIIII!!,2I ' 0000p/ IL

U

OR

_

ItH I

—4 -------- --i----

L) to .44 CQ

ao -

'ex 94nB.IadwaL

CD

0

0

gr

-4

* 0 C,

43

-4

0 C,

D

C)

NACA TN No. 1587

21

43

0 00

0 01f4

C-

0 I0 00

.MCS3

a 00 4.)

0

43

-4 9443

00 -411

ad

0.

4I 00 4o

43

0 43 I.. 04cZ

4 Id C3 LO Gs 0

Cd

'-4 o ,o

a 0'd

H ..-1 0 ,0430

004 00

0 H 0 .rl 0

04 00

04-4 4 04

o4 C.) 0

'do I c:'ti cd .1

N H

0.4

: 0 4 04

Ill

4 -4

IOU 10 t- ------—f------ - $ --- - -, - - -

C) I ci

— it4,

-ci 0

:as

-t --

--

_

\_ --1--

- -t-.1

10 N

io 'T' Ganqviedwel

co

to

S

-4, CQ C)

ci 43 4,

'11

4, C) ci

4, 4,

N0 r. 0

10

22

NACA TN No. $587

'-I -.4

ci

I! 'dr-I

ci 0. 0

I4 oc.

434)

0,0 '-I

4) P.

4) 4,0

'd4

4,0 4.)

4, r1 0

cd

Cd

co

ID

cd

c-I 4)

43 cd (1)

•5•1

CD

U) ••'

'dO

043

I.

• rI0

H CD OU)

S-I

6C

0a

a) U)

a)

4)

a)

4C

2C

NACA TN No. 1587

23

No sprays

Icing

Heating rate (Btu/(hr)(blade))

22,000, 28,000

22

18,000

000

20 40 60

Blade radius, percent

Figure 9.— Radial surface—temperature distribution at blade leading edge. Propeller speed, 1050 rpm.

L

- 1z

- _

co 43

—•00 s-Ia) 0 0 1'

- ' c-----

IIIIIII±\\II -

Bulpvwl

-------

LO 0

-U,

LI LI LI LI 0 CD 10 CQ '-4

'o leaw9.ZTt40).z9dWQJ1

) -4

a

a

0

0 or-I

'-4

MO C) C as 4) a)

a)

as '4

0)

V 03

'-4

0

0

24

NACA TN No. 1587

'SOQ. '-I

c_-I V

o JL4 as -4 ca 14

43 Q..0 '-I 4)3-)

as o.C)

4) a) C P.O VI. 00 14 LO o 0

it) Cl' 0 e

0430

d0 CC * 000

'.4 C) 43

* .00

ioas

0) .0

'i 0 r 0CC

'-4 4) -4 43

V Cd 94 as EQ. 043

'a) a) C) . C) ..a) co '4

as 1-I

00

0 E

0 0) 0 Q. 0)03 0

V u)

co P.r-I E.00 * 0 O'ti o'tio o

a) *0 I CWQ. • 0 C a)

0.,-i ,-44-' ...i '- a) ID a) 'ti r-4

as-I as -4

t1Q,0 0. .1-4 ril

0 -4

.-

0 o 0 C13

L;' C -

C

'-4 .4.) ••

p44.)

0

C 43

.j3 C

o o ID 0

J.

NACA TN No. 1587

25