Icing and De-Icing of a Propeller with Internal Electric Blade Heaters

NATIONAL ADVISORY COMMITTEE

FOR AERONAUTICS

TECHNICAL NOTE No. 1691 ICING AND DE-ICING OF A PROPELLER WITH INTERNAL ELECTRIC BLADE HEATERS By James P. Lewis and Howard C. Stevens, Jr.

Flight Propulsion Research Laboratory Cleveland, Ohio Washington August 1948

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS TECHNICAL NOTE NO. 1691 ICING AND DE-ICING OF A PROP ?T .T.RR WITH INTERNAL ELECTRIC BLADE BET By James P. Lewis and Howard. C. Stevens, Jr.

An investigation has been made in the N&CA Cleveland icing research tunnel to determine the de-icing effectiveness of an experimental configuration of an Internal electric propeller-blade heater. Two atmospheric Icing conditions and two propeller operat- ing conditions were Investigated, In experiments with unheated blades and with heat applied to the blades both continuously and cyclically.

to show the effect of propeller speed., ambient- Data are presented air temperature and liquid-water concentration, and the duration of the heat-on and cycle times on the power requirements and de-Icing performance of the blade heaters.

The extent of ice-covered area on the blades for various icing ax4 operating conditions has been determined.. The largest iced area was obtained at the higher ambient-air temperatures and at low propeller speed. The ohord.wise extent of Icing In practically every case was greater than that covered by blade heaters.

Adequate de-icing in the heated area with continuous appli- the power available but a maximua cation of heat was obtained with power, input of 1250 watts per blade was insufficient for cyclic de-Icing for the range of conditions Investigated.. Blade-surface 0.20 to 0.70 F per second were obtained temperature rates of rise of andthe minimum cooling period for cyclic de-icing was found. to be approximately 91 times the heating period.

INTROIXTION Several methods of obtaining Icing protection for propellers have been proposed including the use of alcohol or other freezing- point depressants , the external electric blade heaters (reference 1), .

amd the passage of a hot gas through a hollow blade (reference 2).

NCA TN No. 1691 • The recent development of hollow steel propeller blades con- structed with a thin-sheet-metal outer akin has permitted the instal- lation of an electric heater within the blade. Such a heater instal- lation does not impair the aerodynamic performance of the blade and aY%1ys1a (reference 3) is free frcn abrasion da mage. Theoretical and flight investigations (reference 4) of heat requirements for ice prevention with continuous heating have indicated, that the electric paver needed. is excessive for current aircraft. An investigation concerned with both continuous and cyclic heating systems was there- fore conducted at the NPCA Cleveland laboratory to determine the icing protection proTided by an experimental configuration of an internal electric propeller-blade heater and the effects of several icing, heating, and propeller operating conditions on the heater performance.

The investigation was made under simulated Icing conditions in the NACA Cleveland Icing research tunnel. Two propeller operating conditions and two icing conditions were investigated, together with several power inputs and heat-on and cycle times. Data are presented to show the effect of propeller speed., ambient-air tem- perature and liquid-water concentration, heating power Input, and the duration of the heat-on and heat-off times on the blade-heater power requirements and performance.

APPARATUS AND The propeller on which the Investigation was conducted was mounted on a modified airplane fuselage locatet' in the diffuser section of the Icing research tunnel. The location of the instal- lation and of the water sprays, together with details of the setup, are shown in figure 1(a); a photoaph of the propeller Instal- lation is shown in figure 1(b). The experimental propeller was a 15-foot, four-blade, hollow, steel propeller reduced to an 11-foot diameter to permit installation in the tunnel. The blade con- struction consisted of a tubular main spar to which steel sheets of 0.037-inch thickness were attached to form the blade profile, an NACA 16-series airfoil section. A semibard sponge rubber within the forward blade cavity served as a vibration dampener.

The blade-form characteristics (thickness-chord, ratio h/b, chord.- diameter ratio b/D, leading-edge-chord radius ratio are given blade angle P ., and design-lift coefficient CL) in figure 2. Electric paver for the blade heaters installed in each of the four blades was supplied through autotranaformera

NACA TN No. 1691 was conducted to the blades by a slip- and an electronic timer and ring assembly mounted at the rear of the propeller hub. Details of in reference 1.

given the power-supply system are Propeller-blade heaters. - The blade heater consisted of a varied distribution of electric resistance wires sewed between two leading- Were of nylon fabric, cemented to the inner surface of the edge cavity, and held in place by the sponge filler. Details of the leading- heater construction and installation including the metallic heated area extended from the edge fillet are given in figure 3. The leading edge to approximately 20 percent of chord on both the thrust and was 47 inches long and extended and camber faces. The heating element 5hjln Inboard from the blade tip to approximately 4 inches from the end of the blade. The total heated area per blade was approximately.

200 square inches. The power density at the inner surface was approximately twice as great at the leading edge as at the rear of - the heatéri. The design power-density distribution on the inner sur- face of the blade for a power input of 1000 watts per blade at five radial stations is shown in figure 4. These curves were obtained from Remilton-Standard Propellers Di'ision, United Aircraft Corpora- 'tion. Temperature limitations of the nylon and the sponge rubber restricted the safe power input to 1250 watts per blade.

Instrumentation. - Instrumentation was provided to measure ambient-air temperature, power to the heating the propeller speed, elements, heat-on and cycle times, tunnel airspeed, and blade- surface temperatures.

The propeller speed was measured by a standard aircraft tacho- meter with an accuracy of ±3 percent. The ambient-air temperature was measured by two thermocouples mounted on . the tunnel turning vanes at each side of the setup and approximately 15 feet downstream of measure- the propeller. The accuracy of the ambient-air temperature F. Blade-surface temperatures were ments was approximately ±3 resistance gages, which temé±ate-sensitiveeleOtriO measured by were similar to strain gages. A total of 40 gages were installed on two adjacent blades. A full description of the surface-temperature- in reference 1. A shift In the calibration to given measuring system galvanometer necessitated the appli- of the blade-surface-temperature of a correction to the indicated surface temperature. A cation correction was obtained by taking the difference between the indicated ambient-air tem- blade-surface temperature and the sum of the tunnel perature and the local kinetic temperature rise. This correction was computed for each gage and each experiment for the dry unheated condition and was then subtracted from the subsequent indicated tem- peratures obtained for each experiment during icing and heating.

NACA TN No. 1691 Observations and photographs of the propeller in operation were made using a stroboscopic lighting system consisting of four flash lamps synchronized with the propeller.

Water-spray system. - Icing conditions were simulated by spray- ing water into the refrigerated, tunnel air stream. A group of 46' air- atomizing nozzles, enclosed in a steam-heated fairing and' located around the periphery of the tunnel innned.iately ahead of the con- traction section (fig. 1(a)), discharged the spray water perpendicular to the air stream. The icing conditions in the tunnel were determined by the method described in reference 1.

CONDITIONS AND PROCEDURE Data were obtained for a series of icing and de-Icing con- ditions each of which was of 10- to 15-minute duration. Propeller speed, ambient-air temperature, and airspeed were maintained con- stant and recorded at 1-minute intervals. Propeller-blade-surface temperatures were recorded either continuously or intermittently.

Photographs were made during operation using the stroboscopic-flash- lamp system. Visual observations of icing and de-icing were made during operation and photographs and sketches were made immediately after each stop.

Operating conditions. - A tunnel airspeed of 120 miles per hour was held constant throughout the investigation. This velocity was the average at the 42-inch-radius station of the propeller as determined by a velocity survey: of the tunnel (reference 1). The propeller was operated at speeds of 800 and 1000 rpm with blade angles of and 30,50, respectively. These propeller speeds correspond to advance diamOter ratios of 1.2 and 0.96, respectively, and were chosen as being representative of low-speed and cruising conditions, respectively. Because of the nature of the tunnel velo- city distribution, the shank end of the blades operated at anl1 i ,negat ve angles of attack; hence, the area of water interception extended further aft on the camber face at this region.

Icing conditions. - The Investigation was conducted at two Icing conditions that were defined by the ambient-air temperature, liquid-water concentration, and droplet diameter. The conditions used were at average ambient-air temperatures of and 180 F with corresponding liquid-water concentrations of 0.3 and 0.7 gram per' cubic meter. An average droplet diameter of 55. microns, as determined by the volume maximum, was founi but no consistent variation with temperature was obtained. The method of calibrating the

NACA TN No. 1691 water sprays and the applicability of the results obtained are dis- cussed in reference 1. The variation of the average liquid-water concentration with ambient-air temperature is shown in figure 5(a) and the radial variation of water concentration is shown in fig- ure 5(b). The values of liquid-water concentration recommended at the Mount Washington Observatory-Weather Bureau meeting of June 1945 and an envelope of values from numerous ground and flight observa- tions as reported by the Army Air Forces in 1946 are also shown in figure 5(a). Subsequent information given in reference 5 shows some values of liquid-water concentration found in cumulus clouds in excess of those shown on the curve. The irked decrease in liquid-water concentration in the tunnel with decreasing ambient- air temperatures (fig. 5(a)) for a constant input of spray water indicates that the amount of frozen-water particles increased with decreasing temperature.

Blade-heating conditions. - Power Inputs of 500, 750, 1000, and 1250 watts per blade were used either continuously or cyclically in the de-icing investigation. The heat-on and the cycle times investi- gated were as follows: Total Cycle Heat-on Heat-off ratio time time cycle time (see) (sec) (see) 15 20 1:4 35 40 1:8 10 20 1:2 30 40 1:4 70 80 1:8 1:4 60 80 140 160 1:8 90 120 1:4 210 240 1:8 80 1:2 40 40 160 1:4 40 120 200 1:4 50 150 The cycle ratio is defined as the ratio of the heat-on time to the total cycle time. AU the heating cycles listed were not investi- the power levels and operating conditions • Two dif- gated at all ferent power inputs or two de-icing cycles were used successively in several of the experiments. In many cases heat was supplied to only one pair of diametrically opposite blades in order that a corn- parison between heated and unheated blades would, clearly show the relative de-icing effectiveness of the heater installation in the same icing conditions.

NACA TN No. 1691 RESULTS AND DISCUSSION The results of the investigation are discussed. with respect to the ice formations on unheated blades, continuous heating, and cyclic de-icing.

Ice formations on unheated. blades. - Views, of typical ice form- ations on the unheated blades are shown in figure 6. The blades are shown .'with the shank towards the bottom and the leading edge at the center. At a propeller speed. of 800 rpm, ambient-air tempera- ture of 17 0 F. a liquid-water concentration of 0.7 gram per cubic meter, and a blade angle of 34.5 0 (fig. 6(a)), a rough irregular rime-ice formation, which extended the full length of the blade on the leading edge, was obtained.. On the camber face the ice extended.

to 100 percent of chord from the shank to approximately 60 percent of the radius and then diminished to the loading edge at the tip.

On the thrust face practically the entire outer 50 percent of the blade was covered.. The negative, angles of attack caused by the nonuniform tunnel-velocity distribution at the inner end. of the blade were responsible for the lack of Ice in this region. At a pro- peller speed. of 800 rpm, an ambient-air temperature of 17 0 F. a liquid-water concentration of 0.7 gram per cubic meter, a blade angle of 30.50 , similar formations to those at a blade angle of 34.50 were obtained on the camber face (fig. 6(a)). Very little ice was obtained on the thrust face because of the decrease in blade angle. Lowering the ambient-air temperature to 2 0 F resulted In a very small formation of smooth fine rime ice (fig. 6(a)). As at very little icing occurred on the thrust the temperature of 17 0 F.

face.

At a propeller speed of 1000 rpm (fig. 6(b)), the ice formations were similar to those obtained at 800 rpm at corresponding ambient- air temperatures and blade angles. Both the chorciwise and the radial extent of icing were less than at 800 rpm. At an ambient- air temperature of 17 0 F and a blade angle of 30.5 0 , a rough irreg- ular rime formation extended to approximately 80 percent of radius.

Light formations on the thrust face were confined to the center por- tion of the blade. At an ambient-air temperature of 22 0 F and a blade angle of 34.50 , the Ice formation on the leading edge extended to approximately 90 percent of radius and. there was indication of runback particularly on the Inner portion of the thrust face.

The radial extent of the ice formations on unheated. blades for the conditions Investigated varied from 50 to 100 percent of the blade radius. A chord'wise coverage on the camber face as

NACA TNNo. 1691 Shank end was obtained with great as 100 percent of chord at the the formations gradually tapering toward the leading edge in a radial direction. On the thrust face a chordwise coverage of 100 per- cent was obtained at the blade tip. The largest formations were obtained at the higher ambient-air temperatures and low propeller speed. The smallest formations were obtained at the lowest ambient- air temperature and low propeller speed. and blade angle. With the exception of the formations at the lowest ambient-air temperature, a rough irregular forination,which showed evidence of natural shedding, was obtained.

The nonuniform nature of the velocity distribution in the tunnel was such as to change the local blade angle of attack, thus causing the position of the ice formation to be different from that obtained in natural icing conditions. As a result, an ice coverage of 100 percent of chord was obtained at the shank end on the camber face and at the tip on the thrust face. In addition, the large droplet size (55 microns) increased the chord.wiae extent of Icing on the camber face over that obtained in natural Icing conditions.

Within the radial extent of Icing, the chordwise coverage In practi- cally every case extended beyond the rear margin of the blade heaters (20 percent of chord).

Continuous heating. - The results of continuous heating of the blades are summarized in table I. The results are given in terms of the percentage de-icing effectiveness, which was defined as the ratio of the percentage of heater area cleared of ice by heating to the per- centage of heater area iced with no heat to the blades. The area iced areas cleared, by heating was taken as the difference between the with and without heat. The heater area was used as a basis for effectiveness because the heaters were ineffective in removing ice beyond the area immediately surrounding them. This method of rating, therefore, does not consider runback and refreezing aft of the rear margin of the heaters. Heating of the blades was continued for - a period of time long-enough to allow the-blade-temperatures--to reach a stable value and until no further improvement in ice removal was observed..

With continuous heating of the blades at a propeller speed of 1000 rpm and an ambient-air temperature of 18 0 F (condition A, table I), a de-icing effectiveness of 100 and 95 percent on the thrust and was obtained with a power input of 500 watts camber faces, respectively, per blade (average power density, 2.5 watts/sq In.) and 100-percent effectiveness on both faces was obtained at a power input of 750 watts In.; per blade (average power density, 3.75 watts/sq condition B).

NACA TN No. 1691 The effect of ambient-air temperature is indicated by the requirement, 10 7 (condition D), of 1000 watts per blade (average at 1000 rpm and -power density, 5 watts/sq in.) for 100-percent effectiveness.

ambient- When the propeller speed was lowered. to 800 rpm at an air temperature of 180 F (condition C), a power input of 1000 watts per blade resulted. in 100-percent effectiveness. At a propeller speed. of 800 rpm and an ambient-air temperature of 7 (condition E), de-ioing effectiveness of 90 and 85 percent for the thrust and camber faces, respectively, were obtained at 1000 watts per blade. An increase of the power input to 1250 watts per blade (average power density, 6.25 watts/sq in.) resulted in 100-percent effectiveness on both the thrust and camber faces (condition F).

The typical rise of blade-stu'face temperatures above ambient- air temperature during icing with continuous beating is shown in figure 7 • The rise of the leading-edge surface temperatures at the 33-percent radius, ambient-air temperatures of 10 and 40 7, and propeller speeds of 800 and 1000 rpm, respectively, for a power input of 1000 watts per blade is shown in figure 7(a). The temperature rise for an unheated blade at 800 rpm is also shown. The tempera- tures for propeller speeds of both 800 and 1000 rpm had. an initial rate of rise for the first 30 seconds of heating of approximately 0.60 F per seeonk, whi3h gradually decreased until a stable value was reached.. At a propeller speed of 800 rpm, a stable value of approxinmtely 67 0 F above ambient-air temperature was attained after about 240 seconds of heating. At a propeller speed of 1000 rpm, a value of approximately 54 0 F above ambient-air temperature was reached in a heat-on time of about 150 seconds.

The rise of the blade-surface temperatures at a point 2 inches from the leading edge on the camber face at 33-percent radius for the sa speed and power conditions and one additional condition is 10 and 40 F, shown in figure 7(b). At ambient-air temperatures of the average rate of rise in the first 30 seconds of heating for both 800 and 1000 rpm was approximately 0.5 0 7 per second. A peak 7 above ambient-air temperature at 800 rpm and stable value of an ambient-air temperature of 4 0 F was reached in approximately 180 seconds and a value of 37 0 7 above ambient-air temperature at 1000 rpm was reached in approximately 150 seconds • At an ambient- air temperature of 180 7 and a propeller speed of 800 rpm, the e initial rat& of rise was considerably lower (0.2 0 F/sea) and peak value of 23 0 7 above ambient-air temperature was obtained in approximately 90 seconds. The difference in blade temperature rise

NACA TN No. 1691 at the different ambient-air temperatures may be attributed mainly to the formation of ice and subsequent release of heat at the lower ambient-air temperatures, which was not apparent at the higher ambient- air temperature. The differences in blade temperatures at the lead- ing edge and 2 inches from the leading edge are primarly due to differences in the local power density. The power density on the

inner surface at the leading edge was approximately 61 watts per

square inch, whereas at 2 inches from the leading edge it was 3 watts per square inch (fig. 4).

Views of typical residual Ice formations after continuous heat- ing are shown in figure 8. Notice should be made of the rather large ice formations at the shank end of the blades. An area at the shank approximately 4 inches long was unheated and the ice forma- tions at this region were extremely difficult to remove. Ice in -the heated area-was also attached to this unheated area, thus increasing the difficulty of ice removal. At a propeller - speed of 800 rpm, ambient-air temperature of 18 0 F. liquid-water concentra- tion of 0.7 gram per cubic meter, blade angle of 30.5 0 , and power input of 1000 watts per blade, indications of runback and refreez- ing are evident. Practically no do-icing behind the heated area was obtained..

A series of stroboscopic photographs taken during icing and de-icing are shown in figure 9. These photographs were made with condition C: propeller speed of 800 rpm, an ambient-air tempera- ture of 18 0 F, and. , a blade angle of 30.5 0 with a power input of watts per blade being applied after 7 minutes of icing.

Do-icing started following the applications of heat and the heated area was almost completely cleared of ice within 5 minutes with very little change In de-icing thereafter.

Cyclic heating. - The results of the investigation of cyclic - heating are-presented in terms of (to-icing effectiveness in table II.

In general, less satisfactory performance was obtained with cyólic than with continuous application of heat. Because of limitations of the heater power input (1250 watts/blade maximum), the require- ments for 100-percent effectiveness could not be determined, for the range of heat-on times investigated. At ambient-air temperatures of 17 0 to 200 F and a propeller speed. of 1000 rpm, the best do- icing effectiveness observed was 90 and 95 percent on the thrust and camber faces, respectively, at heat-on time of 30 seconds, total cycle time of 120 seconds, and power input of 1000 watts per blade (condition K). Decreasing the propeller speed to 800 rpm at approximately the same icing and heating condition (condition Q)

NACA TN No. 1691 decreased the effectiveness to 85 and 90 percent. Although less ice formed on the blade at 800 rpm than at 1000 rpm, the decrease in de-icing effectiveness may be due to the difference in centrifugal force. Lowering the ambient-air temperature from 17 0 to 30 F (condition s) decreased the effectiveness to 5 and 10 percent.

The effect of increasing power input particularly at low tem- peratures is shown by the increase in effectiveness from 20 percent on both faces (condition T) to 60 and 70 percent on the thrust and camber faces, respectively, (condition U) when the power was changed from 1000 to 1250 watts per blade. At higher temperatures of 17 and 200 F and a shorter cycle (conditions M and N), the effectiveness increased from 70 and 75 percent to 80 and 90 percent.

The effect of increasing the heat-on time from 10 to 24 seconds at an approximately constant cycle time is shown by the increase in effectiveness from 70 percent on both faces to 80 percent on the thrust face and 90 percent on the camber face (conditions L and N) when the heat-on time was increased. The de-icing effectiveness was decreased from 85 and 80 percent on the thrust and camber faces, respectively, to 65 percent when the cycle time was increased from 120 to P40 seconds at a constant heat-on time, (conditions P and o).

The effect of varying the heat-on time can be seen by comparing conditions B. I, and K where, for a constant cycle ratio of 1:4, increasing the heat-on time from 5 to 30 seconds gave improved de-icing effectiveness. For each icing, heating, and operating condition the best results were obtained with the longest heat-on time (30 to 50 sec).

In general, the best results with cyclic heating were obtained with a cycle time approximately four times the length of the heat- on time. Cycle ratios as low as 1:2 gave more effective do-icing than the 1:4 ratio only at the highest power input and longer heat- on times.

Typical variations of blade-surface temperatures with time during cyclic do-icing for various icing, heating, and operating conditions are given in figure 10. Results are shown for the tempera- tures of the leading edge at 33-percent radius. The blade- temperature variation at propeller speed of 1000 rpm, heat-on time of 20 seconds, cycle time of 80 seconds, and power input of 750 watts per blade, Is shown in figure 10(a) (condition C). The temperature rose approximately 13 0 F when the water sprays were turned on and reached 33 0 F before the application of heat. Blade heat-on times of 20 seconds resulted in an average rise of approximately F after the start of cyclic heating. The blade temperatures decreased quickly when heating stopped and returned to the freezing point.

NACA TN No. 1691 U The blade temperatures for the heat-on time of 20 seconds and the cycle time of 80 seconds at propeller speed of 800 rpm, two ambient-air temperatures, and power input of 1000 watts per blade are shown in figures 10(b) and 10(c) (conditions M and R,respec- tively). The results at the higher temperature (17 0 F) (fig. 10(b)) are much the same as obtained, at the 750-watt power input (fig. 10(a)).

The blade temperatures again remained above 32 0 F. At a lower ambient-air temperature of -1 0 F, (fig. 10(c)), greater temperature rises were obtained and for most cycles the heated-blade tempera- tures returned to the unheated-blade temperature. The power input and heat-on times were insufficient to raise the blade temperature above freezing and practically no de-icing was obtained anywhere on the blade.

The effect of longer heat-on and cycle times at the same heat- ing and operating conditions is shown in figures 10(d) and 10(e) (conditions S and Q, respectively). Larger temperature rises were obtained with the temperature remaining above 32 0 F for a period sufficient to accomplish de-Icing. As shown in figure 10(d), the heated blade cooled almost to the unheated-blade temperature when heating stopped. The effect of further increases in heat-on and cycle times is shown in figures 10(f) and 10(g) for conditions U and W. respectively.

The blade temperatures obtained at a prcpeller speed of 800 rpm and a power input of 1250 watts per blade are shown in figures 10(h) and 10(i). At an ambient-air temperature of 5 0 F, a heat-on time of 40 seconds, and a cycle time of 80 seconds, average peak temperatures of 37 0 F were obtained (fig. 10(h), condition v). At an ambient- air temperature of 200 F, a heat-on time of 24 seconds, and a cycle time of 84 seconds, the Initial blade rise caused by the water sprays raised the blade temperature to freezing (fig. 10(i),-Con- dition N). The blade temperatures remained above freezing during the subsequent heating cycles with a minimum value of 37 0 F result- ing. The temperature curves indicate too short a cooling period and too much power input.

Views of the residual ice formations on the blades after cyclic de-icing are shown in figure 11. The percentage of de-icing effectiveness obtained for each condition is given in table H.

The photographs show the difficulty in removing Ice from the unheated area at the shank of the blades and the ineffectiveness of the heater In removing ice in the region behind the heated area. Com- parison of figure 11 with figure 8, shows that no significant dif- ference was obtained in runback and refreezing between continuous

NACA TN No. 1691 and cyclic heating. The only significant runback was obtained at 200 F and above (fig. 11(a), 11(c) and ambient-air temperatures of 11(d)). The residual formations for the low-temperature conditions (fig. 11(b)) show ineffective do-icing obtained at law ambient-air.

temperatures. The ice formation on the blades is practically the same as that obtained with the unheated blade (fig. 6(a)).

At a power input of 1000 watts per blade, the blade tempera- ture rate of rise varied from 0.70 F per second at an ambient-air temperature of 30 F to 0.2 0 F per second at an ambient-air tem- perature of 17.50 F. At a power input of 1250 watts per blade, the rate of rise varied from 0.70 F per second at ambient-air tempera- ture of 5 0 F per second at an ambient-air temperature 20 0 F. An average rate of rise of 0.60 F per second was obtained.

The variation in the length of the optimum cooling period with the heating period is shown in figure 12. No significant variation with power input was found for the icing, heating, and operating conditions investigated. The cooling periods shown in figure 12 were the minimum periods in which complete cooling to the temperature at the start of the heating cycle Was attained.

The average minimum cooling period was found to be approximately 21 times the heat-on time.

SUMMARY OF RESULTS - The following results were obtained from an ice-tunnel investi- gation of an internally heated propeller blade at propeller speeds _10 to 200 of 800 and 1000 rpm, ambient-air temperatures from F, liquid-water concentrations from 0.3 to 0.9 gram per cubic meter, heat-on time from 5 to 54 seconds, cycle time from 30 to 240 seconds, blade angles of 30.50 and 34.50 , and power inputs from 750 to 1250 watts per blade: 1. For the conditions investigated, the Ice formations on unheated blades varied from 50 to 100 percent of the blade radius.

A chord.wise coverage on the camber face as great as 100 percent of the chord at the shank end of the blade was obtained with the form- ation tapering toward the leading edge at the tip. On the thrust face a chordwise coverage of 100 percent was obtained at the blade tip. The largest formations were obtained at the higher ambient- air temperatures and at the low propeller speed. The smallest formations were obtained at the lowest ambient-air temperature and at the low propeller speed and blade angle. Within the radial

NACA TN No.. 1691 extent of icing, the chord'tise coverage of icing in practically every case extended beyond the rear margin of the blade heaters..

2. With continuous-heating, 100-percent do-icing effectienese for the heated area as obtained at 18 0 P with a power input of' 750 watts per blade.. At an ambient-sir temperature of 10 p, 1000 watts per blade were required for 100-percent de-icing at 1000 rpm but lowering the propeller speed to 800 rpm and raising the power input to 1250 watts per blade resulted in approximately 100parcent average effectiveness.. Stable maximum temperatures with continuous heating were obtained within periods of approxi- mately 90 to 240 seconds.

3.. De-icing effectiveness with cyclic heating was consider..

ably less than for continuous heating of the blades.. The beet de . iothg (90 to 95 percent) with cyclic heatingas obtained at heat-on time of 30 seconds, total cycle time of 120 seconds, and power input of 1000 watts per blade, for the conditions of propeller speed of 1000 rpm and ambient-air temperature of 17 0 to 190 F. The best de-icing for the conditions investigated required heat-on times of 30 to 50 seconds.. No significant difference as obtained in run- back and refreezing between continuous and cyclic heating..

00 to 17 0 P it is estimated 4.. For ambient-air temperatures of that power inputs of approximately 1500 watts per blade, a heat-on time of approximately 60 seconds, and a total cycle time of ap p roxi- mately 240 seconds would be required for a&equae de-icing. (These values apply only to propellers similar to that investigated.) ]Jo conclusions can be made as to the optimum power distribution as the experimental heater had a fixed power distribution.

The effectiveness of the blade sponge filler as a thermal 5..

insulation was indicated by the small amount of de-icing obtained In the area behind the heater.

0.20 to 0070 P per second 6.. Blade-temperature rates of rise of were obtaInd with the rate of rise a function of ambient-air tem- perature0 A required minim= cooling period of approximately times the heating period as found..

CONCLUSIONS Frca the results of this investigation, several significant conclusions can be drawn concerning the requirements for the design

NACA TN No. 1691 and operation of internal electric propeller-blade heaters. Due consideration must be given to the difference between the simulated and natural icing conditions and to the special construction of the experimental heating element.

1. Despite the marginal performance of the experimental instal- lation, cyclic de-Icing with an internal electric blade heater appears to be a practicable method of propeller ice protection.

2. The chordwise extent of heating on the propeller was inadequate for the icing conditions to which the propeller was subjected; minimum extent of chordwise heating cannot be exactly specified In the absence of data concerning the effects of residual ice formations on propeller performance. Because the results show that the radial extent of icing as varies widely with In propeller speed and icing coalitions, the radial extent of heating will be dependent on the particular design conditions and the tolerable extent of icing on propeller performance. Unheated areas at the shank of the blade seriously impair de-icing and should be eliminated.

Flight Propulsion Research Laboratory, National Advisory Committee for Aeronautics, Cleveland, Ohio, February 4, 1948.

1. Lewis, James P.: De-Icing Effectiveness of External Electric Heaters for Propeller Blades. NPLCA TN No. 1520, 1947.

, Donald R., and Perkins, Porter J.: Investigation of 2. Mulholland .

Effectiveness of Air-Heating a Hollow Steel Propeller for Pro.tection Against Icing. Part I - Unpartitloned Blades NACA TN No. 1586, 1948.

3. Scherrer, Richard: An Analytical Investigation of Thermal-Electric Means of Preventing Ice Formations on a Propeller Blade. NACA ACR No. 4H31, 1944.

4. Scherrer, Richard, and Rodert, Lewis A.: Tests of Thermal-Electric De-Icing Equipment for Propellers. NACA ARR No. 4A20, 1944.

5. Lewis, William: A Flight Investigation of the Meteorological Conditions Conducive to the Formation of Ice on Airplanes.

NACA TN No. 1393, 1947.

NACA TN No.

1691 15 a) U) 0 I I I --O) a) I I — - N — N '- I Na) N 14 I N

zI

a, a, .00 U) 0 0 U) 0 0 0— 0 0) 0 0 co 0 0 a, ciO-i c-i o -4) ') 0 0 0 0 0 0 0 0 0 0 0 0 0 0) 0 •4-4..0 H H H H H a, H •o P.

0 0 0 0 0 04)4-' 0 to 0 0 0 LO U) N 0 0 0 CO H H H H .. 11) U) aS U) LI) U) U) CD 0CJ0 0 0 0 0 ' 4) 10 - 4 1') V) t) ') V) a, — H4' g Cl HICO HIC 4)0,-I 0 U) a) a H I314 4, 43 0I4'-.. N N N CO CO ('0 4-I 0 a, C) 1-4 4-Ia, -I '4) a,4) a,-.

.4 U) U) co H •1I H H H a,4) 4-4 - 0 0 0 0 0 0 0 0 0 0 0 0 0 a) 0 CID a) 00 — H -I H 4. 4 p4 p.4w ci -'-I 4) C) 11 1i C)

NACA TN No. 1691 I III I — III 1.0,4 II 0 I .

I I I I I I I — I - 0 I I I I I I S 00-770q I -'0- I.-4.-1 •4-i.-1to I I I

I7-0 I

P S I I I 'd d 0 I I I-, S s I I I 'I I S - * .-4 .-1 0 -0 I 0 '-I I i-I 0 I I I .-1 i-I I I S - .-1 .-4 .-$ - i-S I .-4 i-0S li-Si-SI I •.-l.-4.-II I 1.00 •0 I 0 P4 IS I I III I' I I I I I I I .-4 i-I i-I S i-S i-I I I I I S I I 4 i-S i-S 14 I i-I 1.00 O0U)0000U)tfl0Il) 000000000000 040 NC)CD0-41-IC'NC)0 tn r4 .0 U)C)(It)C)C)O) Ic) 4-I . 01 'S UI .11) 0 0000 I 000U)U)0 000U)U)U)U)U)0000 S 4-i —414 0 .-1 U) .-4 IS) I C) W dI a) C) C) N N C) CD C) C) cm CD 0) .-4 04-i 0 I IE44-S p, 0 000000000000 00000000000 00U)U)IS)OU)IOOU)0 U)OU)U)U)00000U)0 14 OOr-NNOr-NONO C'J03r-f-00000C.30 043 4' i-Si-S i-S i-S i-S i-Si-li-S .-S.-4i-4i---4.-4.-S I 0 0C30 U) U) mdIi 0 - i-500 00000000000 001I0000000Cb I o ID C) 41 C.) 03 CD a) C) 4I C.) C) U) C) .410) C') U) C') CD CD C) 0) i-Si-Si - i-S 03.-I 03.-Si-S i .-4 C) 4 o — I 0 tflhf)000000000 00.000.4I00000 N 14 03C.304l')C4) H N03ti)t)i)03t4).4IIS) ri 4300 U 0,-I • I 0430 —, NNQr-CDNC)NNC)C) ............

043 w 0 0 0o4) •: i-S 140 NNC)NNC)ONcOOC) 0N00r4N-Sti)U)t1)U).4I WHOW.

fl

1-.

00 00co

P. 0i-S

0.0-

14P.

1:1 Ml.icHS-4 Z0PiS1ZU)E-I.

NACA TN No. 1691 Propeller control 'room Water sprays (a) Location of setup in JoInE research tunnel.

C. 13823 12.7-45 (b) Propeller mounte ii. i ,3-'LnC, research tunnel.

Figure 1. - Propeller-icing research installation.

In Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 19 •' 430

_ . 4 C)

14.4

•____ --

4-4

• ____

.-4 0 ID

1 __

p

I

o bo a) V .

, * a) bo

0 LI -

.

30j i-1 •0

I__I!

S..

.rl

-- N

I --

V

• ___ ___ a)

a) U) a) •rl

__ 1U

C) V

• I, ____ 0

P r.

________ ________ •t1 D

u

• .4

.6

.8

1.0

Propeller-radius ratio, r/R

Figure 2.- Blade-form characteristics for propeller

blade with internal blade heater.

2° Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 Temperature gages ,- Limit of heated area -.--,. .7. --7.

7• --a---- -- ----- <_---._-_j - - -- - .- - -

iI

I OM I - .

(a) Plan form of blade and sectional X-ray views.

C. 20533 1.29.48 Fillet element wires Blade- (b) Cross section of heater at propeller-radius ratio of 0.75.

Figure 3. - Details of heater construction.

21- Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 fiC '-I a, 4) .4.)

Propeller- radius ratio 4.)

.rl C) •tj - Pi Thrust face Camber face 0 _ 3 2 1 0 1 2 3 Distance from leading edge on inner surface, in.

Figure 4.- Design power distribution at 1000 watts per blade (taken from data supplied by manufacturer).

NACA TN No. 1691 3.

—

Envelope of values

I

from flight and

4 ____

/ ground observa-

tions

4)

-

U

Mount Washington

/

- values

C) 4) Icing-research- o

tunnel conditions

Il-I Cr

30 40

.10 0 10 20

Ambient-air temperature, OF

a) Variation of liquid-water concentration with air temperature.

1.

Air temperature bo

(°F)

a

• 0 3

4.) 9

ILI 4.) • C) C)

_ • _ _______

_ C -- A o

0 .2 .4 .6 .8 1.0

Propeller-radius ratio, r/R

(b) Radial distribution of liquid-water concentration.

Figure 5.- Icing conditions for propeller-icing research.

NACA TN No. 169[ Thrust Camber Thrust Camber Thrust Camber Ambient-air temperature, OF Liquid -water concentration, 0.7 0.7 0.2 gram/cu Blade angle, 30.5 34.5 deg 30.5 C. 20535 1.29.48 (a) Propeller speed, 800 rpm.

Figure 6. - Ice formations obtained at various icing and operating conditions with wtheated blades.

Page intentionally left blank Page intentionally left blank

MACA TN No. 1691 27 Thrust Camber Thrust Camber Ambient-air temperature, °F 22 Liquid-water concen- 6.9 tration, gram/cu m 0.7 30.5 34.5 Blade angle, de C-20534 1.29.48 (b) Propeller speed, 1000 rpm.

Figure 6. - Concluded. Views of Ice formations obtained at various Icing and operating con- ditions with unheated blades.

Page intentionally left blank Page intentionally left blank

1691 29 NACA TN No.

LS\

co

of 0)

—

0.

ID 4) D 0) -4 •

-T-t----

-+-

-D-

..-4 ID 4) a,

-

00) I0.

no

-

• 4) ID 4) () -4

r4'd. 0

0 rlO

• p 00. 0 P.O I.00 — - P) 00 to 0 04 - OH D 4) -1 4)0 • 0 U t 0 0) _ • o.o.

------H 0.-I 0 l — _ -1 4) 0)0 0 00 .Tj 0) c-I p. ----jv------- o H

I __

o • a' 'e

---------------'--

• .-I 0).-I 4 i-I4) .00 0 0 c.4.c ;j 0 0) a, 4D 0)0 a) •rI D

P. 0

-- -I- a) rI4) El 0 — a, p.

—

)

It I I, •

0 _

— --_

bo q-f I I - -- -- rI.

-- 0 0 0 0 , 'ojnjodwe .zTe—ueqwe eAoqie esTa exn.zduia

NACATN No. 1691 C) V .1.) .

.4.)

:C) C) .Drl -I C) 544 Cd I Ow to p.) 0 4) 54 4, I C) 0 -1 -ioE 000 0 ø p.000 I 4)0 4Z54 0 CD co I -- $4 54 C) C) p.

.4.)

$1 I C) 4) D —S op.

0 cd 0 C C) 54 to j$4 V C

C) U

VP.

—$43 Cd r4 4) cd C 0 p.

-S S.'

1z -- r4C) D C) cd

---1'---1---

4) I CQ '4 Cd

IIIIIIIi\I

D - -- 0 0

do 'e.xn.zedwe4 ill-luelqmv etoqe asTj 9jnqw9dm94ep-

NACA TN No. 1691 31

I in

In i oa z; –4 4) CO LI CII CD 0 to a ci HJC%JLL) Q C\ • r'., • 0 o co 0 gj tO -@

if

H ID ID 0 4D N cO L O co H 0 0 to H ,-1 as ID ID O 1!) 00 • • 0 0 0 0 0 0 tO H H 4, c-s ci ID C) ID •0 C) ID C) —1 —I ci 0 4' 4' —

'-4 -'-I

4' H

o o '-I H +' H -P c 0 'dO '-I I

rIH ID i

4' ' o 0 .

ci o ID 0 p-1 I1 I-4' pq p1

Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 6 13 5 12 -I 1 8 - Id P.

rxi O7 rd P4 C. 18002 2.26.47 Figure 9. - Icing and de-Icing of leading edge and thrust face with continuous heating.

Condition C (table I): propeller speed, 800 rpm; ambient-air temperature, 18 0 F; blade angle, 30.5°; power input, 1000 watts per blade.

3L( Page intentionally left blank Page intentionally left blank

NACA TN No. 1691

35; SQ V

•0

Q1-4 0 .D .4 0.0 4) .4 Cd •0

4 . 'O V 0

11 a

541-4 .4 4.)

0.•

a4 .1

4)

Q

4)ft

54.1 0

.5 0 U) ,_ V +3 00. 54 .5 .0 • 0.

no a.

NIt) 0

• 4.) 0

CD

tL) V 1

54- 004.

-r4 5443

to aS

•• IC)

F1

0.0

is I-Is

0r4 P

op

•0

00 V

j

VtI

(d)

0.0 0 .-I

00 .40 V .0.4 0.

$

00 0

1-4

r1OD ,-I

-4

0.0 0

o

54.4 0.4 . ) 0 Q —I-I H C) .4 Hp...

C) • .

I-I • • -0 •.-I V 00 U 43 • •5-0 1-4 Is 0.5 -

'en.xeduie ej.zn_epT

di

36 NACA TN No. 1691 'V 0.0 U) 0U) -I, 43 .-1 cd i 0 to o 0.-U) 43 Q• • E 43 13, 0 r. 4) 14 14 -I 0 ,0 aJ4 to IG) aS P0 00. ; .0.. IV aSiz) 0i Id co r .P tion •0 £ ad U) i 00 00 tO VU) aStO 00.

is n 040 -4 IV

V.P

...a aS ai to 0.-I 4) HtIO co

.-

a 0 00 -.4 00 4 00 0 i 00 00 0.0 ..4V 00 4) U) • as p.

4 4l co 00 girl H r4 CQ aS 0

•

0.-H 00 aS c1 0.i H i-I H 0 00 .

H 0 H cd 42 04 4)V - •- 43 o r10 0 • 0 0 I CD 43i) -4 0 0u V : 0 -4 '-P 0 l.

d0 l eanqisaedmeq eo.rns—epB

NACA TN No. 1691 ••rn• -..rnrn.

a a Id

.i• MENNEN

_

•0 C

MEMMEM

I4 C o 0.4 .43 t C .4

MEMEMEM EMEMMEM

_

-4, V .43 o . O4) to $4O Co 1, gu-4 V

ONIONu_U___

• -H

434) C p40.

o .laC

tVAiiUiU

.40 W to

r'

a.

C

MENI M I&A

.40 C-.l a .

- . V Li: o 0

smismU ME PAFA EII

M

OD ce C

4)

•0

C to

E_tPIIIU U_lUiUHhl

-V C C ,-4 ••. t') o

IUU U__i!III

11111

o o.i

h

ID Idol C u, OV C o

mpJ..ii.;.__rn--un

- a C - 00 043 e. a'

; •

CQ 0-0 00 N)V o.a .4V 00 4.1 C 0O C 0.$4.

NEI uiU' ..iuIi

CS CC

-4 _4 0 040 0 0 0..-44 I.

1' 0 c -o 00.

..-4

_

4-4 He

0 r 0 CUV

•atrniiu a!rlr•_

to ,.4) 4 .00 C 43

lmlimmil •

a 0

•

0O 00 S

_I_ __I__ . _

.4 C CD 43 • .4 0 -4

UiIIiUU rnEU uRN

0-V

----I-"--. __

—'-.4 vi •iiiu..

--

0 0 ..

to ao 'eIr4Bzedwe eoBjan.-opg

NACA TN No. 1691 MEMEMEM MEMO MEMEMEEM MEMO MENNEN NONNI MEMNON MENEM (0 MEMNON NONE .

E -NEME I

(0

_

0NOME

CD

MENEM m ummil

•_ m olumn

_l .

.q.o

loll_ o nionII KIUI

Q) E

11111

ME N EM 0 1UolM__

C'.

0)

--!iUllI U_ ii

Cd

mm u s l imm m

II __

C'.

M EMO ii_.Im.i._II

II 11lI'iUII

(0

N o Iimminum

A INO_U

0 UU! II

OD

E m L qu, b q h !-m HhI1UI

111= 1^ 4 14mom M m'-mm

1= a=

momm mimumms II

ill

-Jo

L )

0 0 0 0 (0 q( C'..

d0 'e.xniedwe e3JartB_epI

NACA TN No. 1691 39 I • 4

MEMEMMEMEMEME

0coV .00

MOMMEMEMEMEME

-4 04) 00.

o.

CQO 4)

MOMMEMMEMEMEMI

4) -

MEMMEMEMEMEMEM

4)0 4) 4O I

MENNEN 0. 0

MENNEN

• E 4 .'

• P.

MESIONE MENNEN V4 0

to W) . 0 I

_ V _ _ 4) 0 I-i •

-IP. 0.4 0 10 H -. I 10 V

r IMM 0

'01 OD • ,-14.)

t. •0

fi MENIMME • 00

o • ..0 '00 OP.

to II MENNEN -0 mt') UIU • HI 0. .4 1.0 '04' 0 0 II

MENNEN

uil MOSEME __ 00 I

Orl r4to cO,c *-. 0 0 0-I '00 I OC 00 V 0.0 .4V 004.)

MEMEMEM

Effismimm • 0 00

p.

0 1.,-I 0 0 rx MENEMEM 0 • 0.-.'

00 0 U P.4 .4

MEMEMEM _ ______ .4.)

MRSEME

—. 0 I-I i-I I-I 0 D

MESIMINE __i ______ __ 0 oll

'0 -4 r-4 0 .0•..

4)0 rI -' C) ZO 0

EMEMEMMINNISREM

) 0 I 42 CII MMEMMEMIIIEMRIMMM V o g • 0-I 4.)

EMEEME MO

MM E N SS_

0 .4 I_I _l0 '— .

to d0 'e.zn.zedtue

qO Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 Thrust Camber Thrust Camber Condition (table II) P Ambient-air temperature, 0F 21 Liquid-water concen- 0.8 tration, gram/cu m 0.9 Cycle time, sec 120 C-20537 1.29.48 (a) Propeller speed, 800 rpm; heat-on time, 30 seconds; blade angle, 34.5 0; power input, 750 watts per blade.

Figure 11. - Residual Ice after cyclic do-icing.

Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 43 Thrust Camber Thrust Camber Condition (table II) N R Ambient-air temperature, °F 17 -1 Liquid-water concen- trat ion, eras/cu m 0.7 0.3 C- )538 1.29.48 (b) Propeller speed, 800 rpm; heat-on time, 20 seconds; cycle time, 80 seconds; blade angle, 34.50 ; power input, 1000 watts per blade.

Figure 11. - Continued. Residual ice after cyclic de-icing.

(]L/ Page intentionally left blank Page intentionally left blank

No. 169 45 NACA TN Thrust Camber Thrust Camber V Condition (table II) N Ambient-air temperature, °F 20 Liquid--water concen- 0.3 0.8 tration, gram/cu m Heat-on time, sec go Cycle time, sec C-20539 1.29.48 ) Propeller speed, 800 rpm; blade angle, 34.5; power input, 1250 watts per blade.

( c Figure 11. - Continued. Residual ice after cyclic de-icing.

Page intentionally left blank Page intentionally left blank

MACA TN No. 1691 47 Thrust eber Thrust Camber G Condition (table II) H Ambient-air temperature, °F 17 Liquid-water concen- 0.8 tration, gram/cu w 0.7 Cycle time, sec 80 160 w C-20540 1.29.48 (d) Propeller speed, 1000 rpm; beat-on time, 20 seconds; blade angle, 30.5 0 ; power input, 750 watts per blade.

Figure 11. - Continued. Residual ice after cyclic de-icing.

ILA Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 Thrust Camber Thrust Cazooer Condition (table II) B I Ambient-air temperature, °F 17 Liquid-water concen- tration, gram/cu m 0.7 0.7 Heat-on time, sec 5 20 Cycle time, sec 20 80 ACA C-20541 1-29-48 (e) Propeller speed, 1000 rpm; blade angle, 30.50; power input, 1000 watts per blade.

Figure 11. - Concluded. Residual ice after cyclic de-icing.

Page intentionally left blank Page intentionally left blank

NACA TN No. 1691 51 ].4C 12C 1OC

()ZI

ID 8C Heater power bo r.

input •14 6( (watts /blade)

/ 0

0 1000

/

- 4(

0/

^z

50 40 50 10 20 Heat-on time, see Figure 12.- Variation of optimum cooling time with heat-on time.