eet NACA TN No. 1472 16 JAN 1948 NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS ‘TECHNICAL NOTE No, 1472 4 ‘THE CALCULATION OF THE HEAT REQUIRED FOR WING THERMAL ICE PREVENTION IN SPECIFIED ICING CONDITIONS By Carr B, Neel, Jr., Norman R, Bergrun, David Jukoff, and Bernard A, Schlaft Ames Aeronautical Laboratory ‘Moffett Field, Calif. SEE ‘Washington, December 1947 FOR REFERENCE o WOT TO BE TAKEN FROM THIS ROOTNATIONAL ADVISORY COMMITTEE FOR AERONAUTICS TECHNICAL NOTE NO. 1472 THE CALCULATION OF THE HEA REQUIFED FOR WING THERMAL TOR FREVERILON IV SPEOLFIED IOING CONDITIONS By Carr B. Neel, Jr., Norman R. Bererun, David Jukoff, and Bernard A, Sohlart SUMARY As a result of a funfamental investigation of the mtecrological. conditions conducive to the formation of ice on aircraft end e study of the process of airfoil thermel ice prevention, previously derived equations for calculating the rate of heat trensfer from airfoils in doing conditions were verified. Knowledge of the manner in vhich water is deposited on and evaporated from the surfece of a heated airfoil wes expanded sufficiently to allow reasonably accurate caleu- lations of airfoil heat requirenents, The research consisted of flight tests in naturel—tcing conditions with two 8-foot-chora heated airfoils of different sections. Measurements of the meteorological variables conducive to ice formation were made simitensously with the Procurement of airfoil therml date, It wes concluded that the extent of Imowledge on the meteorology of 4oing, the impingement of water drops on airfoil suwfeces, and the Processes of heat transfer end evaporstion from a wetted airfoil surface has been increased to a point where the design of heated wings one fundamental, wet-air basis now can be underteken with reasonable certainty. ‘TRADUCTION For © period of several yeere, the National Advisory Comittes for Aeronautics” has conducted research on the prevention of ice formtion on efreraft through the use of heat; During this time, research of a fundamental nature on the problem of thermal ice prevention was retarded by the more urgent eed for development of ice-prevention systems for specific airplanes in military servicea Satisfactory ving- and tail—curface thermal, ice—prevention systems for a Lockheed 12-A, Consolidated B2h, Boeing B17, and2 NACA TN No. 1872 Curtise-Hright 0-46 atrplenes (references 1, 2, 3, and 4, respectively) were designed, febriceted, and tested in netural—tcing conditions. Windshield thermal ice-prevention systems vhich proved edequate in the doing conditions encountered were provided for the 12-A, Boh, end O-6 airplanes. Eech ving- and tail-curface design was based on establishing, for flight in cleareir conditions, e surface— temperature rise above free—stream temperature vhich experience in eimlated— and naturel—icing conditions had ehovn to be adequate for doe prevention, This empirical method, vhile proving satisfactory for the airplanes tested, ves limited, since 1t ves not established on @ fundamental besis, and s more basic rrocedure founded on designing for the conditions existing in icing clouds vas needed. ‘The NAGA at present 16 engaged in en investigation to provide a fundamental understanding of the process of thermal ice prevention in order (1) to establish a basis for the extrepolation of present Limited asta on heat requirements +o metecrclogicel and flight condi~ tions for which test data are not availeble, (2) to provide data for improving the efficiency of thermal 1ce—prevention equipment, end (3) to provide a wet-air, or meteorological, basis for the preparation of design specifications for thermal ice—prevention equipment. The research consists of an investigation of the meteorological factors conducive to icing, and a study of tho heat—transfer processes vhich govern the operation of thermal ice-prevention equipment for airfoils and for windshielé configurations. The airfoil heat~transfer phase of this investigation consisted of the measurement of the factors affecting the transfer of heat from airfoil surfaces during flight in natural—tcing conditions ‘Those data are correlated with the simtaneous measurements of the metecrological paremters vhich influence the heat—transfer proce: and are analyzed for the purpose of establishing a wet-eir ico Prevention design basis for airfoils. The first approach to the toing heat—trensfer problem on 8 fundemontel basis was made in England by Hardy and Mann prior to gle. In thie study, @ method for the calculation of heat transfer from a heated surface subjected to icing conditions was presented and substantiated by measurements in an icing tummel. eter work by Hardy in which these heat—transfer equations were modified for general application 1s presented in references 5 and 6, Reference 5 contains information on the protection of all aircraft components against tco accretion. Reference 6, prepared during a period of active participation by Mr. Hardy in the NACA icing research program, presents an analysis of the dissipation of heat in conditions of ieing from e section of the heated wing of the C6 airplane (reference 4).NACA TH No. 272 Other research in the present NACA investigation hes been reported in references 7, 8, and 9. Reference 7 gives the first measurements in this program of the liquid-water concentration in clouds. References 8 ani 9 deal with the meteorological aspects of icing conditions in stratus clouds end in precipitation areas of the verm-front type. Research on the problem of heat transfer from airfoils in condi— ‘tions of icing has also been conducted by other laboratories. In reference 10, the transfer of heat from surfaces subjected to icing conditions on Mount Washington has been studied. The General Electric Research Laboratory has conducted a number of investigations on this phase of icing. A sumary of this work end e list of reports 1a presented in reference 11. A comprehensive report by the Army Air Forces on the development and epplication of heated wings is contained in reference 12. _ In continuation of the present icing progren, the O-N6 airplane was equipped with special meteorological and electrically heated test epperetus, and flovn in naturel-icing conditions during the vinters of 1915-46 and 19h6-47, Flight tests were conducted minly along airline routes over most of the United States, ‘The meteorological, data recorded dwring the icing conditions encountered in the two seasons are presented and discussed in references 13 and 1h. 5 report presents an analysis of the data obtained during the aisle ant IghGby winter coasons with two electrically heated alr foil sections, The data vere analyzed using the heat—transfer equations developed by Hardy. (See references 5 ani 6.) A considera ‘tion of the area and rate of water impingement on one of the airfoil sections based on an analytical study of waterdrop trajectories (reference 15) is also presented. An attempt is mde to further the Knowledge of the process of airfoil thermal ice-prevention, ‘The appreciation of the NACA is extended to United Air Lines, Inc., the United States Weather Bureau, and to the Air Materiel Command of the Army Air Forces for aid and cooperation in the research, In particular, the services of Major Jemes L. Murray of the Air Materiel Command, Army Air Forces, and Ceptein Cerl M. Christenson end First Officer Lyle W. Reynolds of United Air Lines, who served as pilots of the research airplane, were a valuable aid to the conduct of ‘the investigation.op Py » ACA TW No. 1472 srmais The following nomenclature is used throughout this report: radius of vater drop, foot airfoil chord length, foot specific heat of air at constant pressure, Btu per pound, degree Fahrenheit specific heat of water at constant pressure, equal to 1 Btu Per pound, degree Fahrenheit concentration factor, defined in equation (6), dimensionless saturation vapor pressure with respect to water, millimeters of mercury ‘vater-drop collection efficiency, defined in equation (10) eoveleration of gravity, equal to 32.2 feet per seconi, second convective surface heat—transfer cosfficient, Btu per hour, square foot, degree Fahrenheit total surface heat—transfer coofficient, Btu per hour, square foot, degree Fahrenheit mechanical equivalent of heat, equal to 778 foot-pounds per Ben ‘thermal conductivity, Btu per second, square foot, degreé Fahrenheit per foot @imensionless drop-inertia quantity, defined in equation (5) latent heat of vaporization of water at surface temperature, ‘Btu per pound iguid-water concentration of icing cloud, pounds of water per cubic foot of air weight rate of waterdrop impingement per unit of surface area, pounds per hour, square foot weight rate of water floy aft of area of water-drop impings— nent per foot of span for one side of airfoil, pounds per hour, footNAGA TH No. 1872 Ve Yo e/a weight rate of waterdrop impingstent per foot of span for ons side of airfoil, pounds per hour, foot concentration of liquid water contained in drops of each size ine drop-eize distribution, pounds of water per cubic foot of air vercmotric pressure, millimeters of mercury Prandtl number (cpu/i), dimensionless unit rate of heat flow, Btu per hour, square foot Reynolds number for airfoil (Voy/), dimensionless free-strean Reynolds number of water drop relative to speed of airfoil (2Vay/u), dimensionless @istance measured chordwise along airfoil surface from stagna— ‘tion point, feet ‘tomperature, degrese Fahrenheit local velocity just outside boundary layer, feet per second free-stream velocity, feet per second weight rete of eveporation of vater per unit of surface area, pounds per hour, square foot weight rate of evaporation of water per foot of span for one side of airfoil, pounds per hour, foot Gistence measured chorawise along airfoil chord line from zero-percent chord point, feet evaporation factor, defined in equation (22), dimneionless airfotl ordinate, fect starting distance of water drop above projected chord line of airfoil, fest pressure altitude, fect ratio of saturated to dry adiabatic lapse rates exponent of Prendtl nuber, 1/2 for leminar flow, 1/3 for ‘turbulent flow6 NACA TW No. 1472 7 specific weight of air, pounds per cubic foot tw specific weight of water, equal to 62.4 pounds per cube foot B viscosity of air, pounds per second, foot Subscripts 2 refers to conditions at edge of boundary layer x kinetic ° refers to free-stream conditions me mean effective refers to conditions at airfoil surface sL sea level ANALYSIS During flight in foing conditions a heated wing is cooled by convective heat transfer, by evaporation of the water on “he surface, and, in the region of droplet interception, by the water striking ‘the wing.? ‘The rate at which heat mst be supplied in order to maintain the wing surface at a specified temperature is, therefore, a fuotion of the rates of convection, evaporation, and water impingement. Equations for expressing this heat requirement are presented in references 5 ani 6, These equations, with slight modification, are used throughout thin report. Expressed 2s an equation, the mit heat loss q froma partially or completely wetted surface éxposed to icing conditions my be stated: a= Qw + do + do Q@) where ww heat loss due to warming the intercepted water w heat loss due to forced convection % heat lose due to evaporation of the impinging water Each of these individual heat flows will be analyzed. me heat loss due to vediation 1s amall and can be neglected.« NACA TN No. 1572 Heat Loss Dus to Warming the Intercepted Water In the region where water droplets strike the wing, the heat required per unit ares to heat the water to surface temperature is wo [te (to m)] o The term Athy is the kinetic temperature rise of the water caused dy stoppage of tho droplets as they strike the wing. The value of tn, 18 given by Athy = eae @) where V ie the free—styeam velocity in feet per second, The value of Ati, is less than 2° Fahrenheit for airplane speeds up to 200 miles af hour end, for the calculations presented in this report, the term has been neglected. Equation (2) thus becomes: ay = Ma (testo) (4) ‘The weight rate of vater impingement on the wing, the area of impingement, end the distribution of the water over that area are important factors in the heat-transfer analysis. In addition to the effect of the amount of water intercepted on tho value of qy in equation (4), the evaluation of Mz provides an indication of the quentity of vater vhich mst be maintained in a liguid state until it either evaporates or rms off the trailing edge if the formation of ice aft of the area of impingement, normally termed "runback,” is to be avoided. ‘The area of impingement influences the extent of heated region to be provided at the leading edge, while knowledge of ‘the @istribution of water impingement is required in the calculation of the heating requirement in areas where water is striking. Calowlations have been made by Glauert (reference 16) for the trajectories of water drops about cylinders and an airfoil. In thie work the assumption was made that the drops obeyed Stckes' law of resistance. At the speeds of flight, hovever, Stokes’ law no longer strictly holds, and Langmir and Blodgett (reference 17) computed = series of drop trajectories about cylinders, spheres, and ribbons, taking into consideration deviations from Stokes' lav. These compi- tations were undertaken on the assumption that the trajectories for8 MAGA TN No. 1472 gylindere would apply to airfoils 1f the airfoil were replaced by an “equivalent” cylinder (reference 12). . . - Preliminery calculations based on references 16 and 17 indicated that, for lerge values of drop size and etrepeed, the assumption of the guivalent cylintor vould not hold for airfoils. Therefore, more extensive calculations were undertaken to determine the drop trajec— tories for one of the test airfoils of this research, an NACA O02 sirfotl at O° angle of attack. In these calculations, presented in detail in reference 15, a Joukewski airfoil (the contour of which closely approximates that of the NACA 0012) was used to supply the stream lines since the Joukowsk! stroan lines and volocity fel can be computed with relative ease. The basic equations presented in reference 16 wore used with modifications for deviation from Stokes* lav as given in reference 17. ‘Tho procedure followed was to start a given distance forward of the airfoil and calculate the paths of the Grops using a step-by-step integration process. Results of these computations ere presented in figure 1. The curves shomn establich the distance s, measured from the stagnation point, at which a given drop will strike the eirfotl vien starting a distance yo above ‘the projected chord line. Curves are presented for various velues of K, where *-§ OCH) 2 It should be noted that the curves of figure 1 apply strictly only for a Grop Reynolds number Ry of 95.65, that is, only for pertio— ular coubinations of drop size, airspeed, altitude, end air tenpere— ture. The value of 95.65 was chosen as being the Reynolds number corresponding to average conditions of drop size, airsyesd, altitude, and air temperature experienced during tho tests of this investiga— tion, However, the curves of figure 1 can be used for a range of Reynolds numbers on either side of 95.65 without serious error. Due to wracticel considerations, these curves wore used in the enalysis of the date presented in this report, even though the Reynolds number differed somewhat for every case. Area of vater omont.— ‘The end points of the curves shown tn £59 TE UOSTS TS GEESE Location at which Srope of @ partion Wer K value will strike the airfoil. Beyond this value of s/c no drops of this K value will hit, Thus, the broken line in figure 2 establishes the aree of impingement for all values of K. Rate of water impingement. The rate of water impingement at a specified point on an airfoil is a function of the area of impingement, ‘the velocity of flight, the liquidvater concentration of the airWAGA TH Wo. 1472 9 stream, and the distribution of the intercepted water over the surface. ‘This latter factor, called the concentration factor © is represented by the ratio of y, to 8, or: cate For point values, ox on or tore exactly, 830 ons (6) ‘The veight rete of water impingement per unit of surface area in pounds per hour, square foot, then, is, My = 36007 mC Mm ‘Tt 4s apparent from equation (6) that © is simply the slope of ‘the curves shown in figure 1, A plot of the mesured slopes of these curves asa function of s/o is presented in figure 2. Using values ef © obtained from figure 2, the weight rate of water impingement at any point on the surface can be caloulatec from equation (7). Tn the case of a cloud, where the water drops are not of uniform size, but instead follow a pattern of size distribution, the rate of impingement can be computed if the dictribution is known or assumed. ‘The rate of water impingement at eny point is the sum of all the rates of impingement of the volume of water contained in each drop size. ‘Equation (7) then becoms M, = 3600 v) 20 @) Where n is the concentration of ligula water contained in drops of @ particular size ani O is the concentration factor for the EK value corresponding to thet drop size. ; In order to esteblish the possibility of runbeck forming att of ‘the heated area of a wing, 4t 1s necessary to know the totel quantity of water intercepted per unit of wing span, ‘This rate of impingement, @encted as Mg in pounds per hour, foot span, is given by10 NACA TN No. 1472 Mg = Ma ds (9) f A more rapid method for the evaluation of My utilizes a ourve of collection efficiency EZ asa function of E (fig. 3). Collection efficiency is defined as iz a x Tuas oo hore You1mt is the value of yo for which drops of = particular KE value just mies the airfoil, and Ymox is the maximm ordinate of the etrfoll. The equation for computing Ms, then, 1s Mg = 3600 EVm ymax qn) Using figure 3, the rate of water impingement can be computed for each of the drop sizes in the assumed or measured drop-size distri— Dution, The totel rate of impingement 1s the sumation of these individual rates Heat Loss Due to Forced Convection ‘The unit heat flow from the surface of @ body in an air stream resulting from convective heat trensfer can be expresse’ Go = b (toto) (12) Where tg is the surface temperature and toy is the kinetic tempor ature of the free-stream air at the point for which the heat flow is being computed. The fector h is the convective heat—transfer coefficient and may be evaluated by measurements in clear air or by calculation using the methods presented in references 18 and 19. valuation of the term to, will now be discussed. ‘The surface of an unheated wing moving through the air will assume a temperature somevhat higher than that of the free air stream Decause of stoppage of the air particles in the boundary layer next to the surface. This temperature rise is of importance in the calcu- lation of heat requirements for ice prevention in that it establishesACA TN No. 1472 an ‘the datum point from yhich the temperature of the surface mst be raised to obtein the desired temperature, tg. Tho value of the temperature rise in clear eir, from equations derived in reference 5, is, for leminer flow, -Seb-RG-)] ow and for turbulent flow, onze [2-8 G-o)] aw where U 4s the local velocity Just outside the boundary layer at the point along the surface vhere the value of Ot, is being calculated. In clouds, the kinetic temperature rise is reduced, dus to eveporetion of water from the surface. Assuming the surface is completely wetted with water, the velus of the temperature rise for laminar flow becomes og 8 G-=)]- Batey where er = 00 # (26) end co, i the vapor pressure at saturation at the wet kinetic tem perature, toe The value of to, 18 toy = to + Btyae WACA TH No. 1472 toy = to + ies 2-8 (2-2) ] - ovceate Ce) (a7) Equation (17) 4s for laminar flow, ‘The equation for turbulent flow 4s the cane, but vith the exponent of Pr changed to 1/3. Tt can be seon that thie equation mst be solved by trial, eince the value of Coy 48 dependent upon the temperature toy. ‘Experiments in clouds, in the process of calibreting a free-air ‘thermometer installation (reference 13), showed thet by mltiplying the clearetr kinotic~temperature rise by the ratio of the saturated to the dry adiabatic lapee rates, good agreement between the values of Kinetic temperature rise calculated in this manner end the measured values vas obtained. Since use of the ratio of the adiabatic lapse retes vas substantiated experimentally, and since equation (17) mst be solved by trial, a somvhat laborious procedure, the following equations were used in this repart to calculate values of toy: For laminar flow, warts [8 OD] and for turbulent flow, tog = to + TE [b-8 @-F)] (as) Values of ae/ag, the ratio of the wet~ to the dry-ediabatic lapse rates, are obtained from figure lk, The uso of the lapso-rate ratio tn equations (18) and (19) is somi-cupirical. he limitations of this simplification in the calculation of Kinetic-temperature rise of airfoil surfaces in clouds are not kmovn, Belov speeds of 200 mils per hour, however, these wquations cen be used with emall errer, eince ‘the kineticntemporeture rise is lov. (a8) alg Heat Loss Due to Evaporation of the Water on the Surface ‘The amount of heat removed from a wetted surface as a result of ‘the evaporation of water on that surface can be expressed: de = Lp We (20)HAGA TN No. 1472 33 From reference 6 the relation between g_ ani the convective heat-transfer coofficient h can be expressed for a completely wetted surface as: de = b (K-12) (teto,) (a) where ty (tes rans (Se) ee (ea) By substituting average veluss for Ig and cp, equation (22) can be rewritten, r=14+3.5 (=) aE (23) fhe velues chosgn for Ig and cp are 1100 Btu per pound and 0.24 Btu per pound, mheit, respectively. ‘The factor Pst/P: is the ratio of the standard sea-level pressure to the local static pressure. Tt should be noted that the evaporation factor X applies only when the surface is completely wetted. If only partial wetness yrevails, the value of X must be modified according to the degree of wetness. Totel Heat Loss froma Wetted Surface Sumerizing the heat losses due to vater impingement, convection, ‘and evaporation, equation (1) can be written: 2 = Me (toto) + (tetoy) +h (X—2) (tertox) Which reduces to @ = Ma (te-to) +h X (te~tox) (2h)uy WACA IN No, 1472 Aft of the region of water impingement, M, = 0 and equation (2h) becomes a= bX (tetoy) (25) DESCRIPTION OF EQUIPMENT 411 tests reported herein were made in the O-K6 airplane shown 4n figure 5. ‘he airplane had been modified to provide thermal ioo— Frevention equirment for vings, empennage, windshield, and propellers. A description of the thermsl system for the wings and empennsge 1s given in reference 20, The windshield system was altered for the flights as described in reforence 21, Protection for the propellers ‘vas provided by electrically heated blede shoes. The meteorological equipment used during the tests to masure ‘the free-air temperature, liquid-water concentration, drop size and drop-size distribution 1s described in references 13 and 1h. vo electrically heated test airfoils wore used to obtain fundamental data on the process of ving thermal ice prevention. Each airfoil was mounted vertically on top of tho fuselage of the O-46 airplsne, as shown in figure 5. Tho test afrfoil installed during the winter of 1945-6 had an HACA 0012 section. Yor the tests in the winter of 1967 the airfoil had an NACA 65,2-016 section in order to provide test date for low-drag sections, as voll es conventional sections. Both sections are symmetrical, ani the models vere installed with the chord line in tho plane of symmetry of the ain plend; that is, et zero angle of attack for unyawed flight. Ordinates for en NACA O0l2 airfoll are given in reference 22, and for an WACA 65,2-O16 airfoil, in reference 23. igure 6(a) shows the HAGA 0012 airfoil mounted on the fusolage. the NACA 65,2-016 airfoil was mounted as shown in figure 6(b). A clear plastic dlister, shown in figures 6(a) and 6(b), allowed the airfoils to be viewed and photographed in flight. Both airfoils haa an 8-foot chord and e 4.7-foot span, with a faired square tip. A heated test section of 1-foct span vas located 2 feet above the top of the fuselage. It had been determined Previously, by means of a prossure survey, that the test-section ocetion was well above the edge of the fuselage boundary layer. Electrically heated guard sections vere built around the leading~ edge region on both sides of the test section for the purpose of preventing any disturbance of the sir floving over the test section Which might have been caused by 1ce accretions in the region of the guard sections.MAGA TH No, 1472 6 NACA 65,2-O16 Airfoil Model Construction details of the 65,2016 airfoil motel ere shown in figure 7. The mtel portion of the structure consisted of eluxinm riba ani ekin supported from the fuselage by tvo spars. ‘The test section vas mde up of a 3/8-inch-thick plastic bese and e sheot of plasticimpregnated febric, 1/éi-inch-thick, on top of which 1/2-inch Wide, 0,002-inch-thick, olectrical resistance heating strips vere cemented in a spanwise direction spaced 1/32 inch apart. A covering of the 1/6\inch-thick plastio~imprognated sheot vas lai over the Tesigtence strips, end on top of this vas cemonted a skin of 0,006-inch—thiok luminm, Each 1/2-tnch-vide heating strip wae connected to individual Ings located along the edges of the test section, This provided means for chordwise adjustment of the pover @istribution by 1/2inch increments. The heated area of the test section extended beck to 77 percent chord on the left side ani to 17 percent chard on the right aide. ‘The guard sections vere constructed in the sane manner as the test section, with the exception that the aluminum skin was 0.011 inch thick, The heated area of the guard sections extended to 17 percent chord on both sides of the model. Yeasurenents of the temperature of the aluminum surface of the test section were obteined by means of thermocouples. Fine iron constantan thermocouple wire was rolled flat to produce a strip approximately 0,002 inch thick end 1/16 inch wide. These strip ‘thermocouples were laid in epenwise grooves about 3 inches long cut in ‘the aluminum skin. The thermocouple junctions were located in the middle of tho grooves, and the leads pessed through holes et the ends of the grooves into the interior of the model. Aluminum wes sprayed into the groove over the strip thermocouple for a distance of about 3/16 inch on either side of the Junction. Thus, the thermocouple Junction ves bonded to the aluminum skin, alloving accurate surface— temperature measurements to be meade. The remainder of the groove on either side of the aluminum sprey was filled vith e nonelectrically conducting mteriel. Thermocouples were located at the center of the test section at 1-inch chardwise intervals in the leading-edge and calculated transition regions, and at 1-1/2~inch chordwise intervals in other regions. Surface temperatures were recorded by means of self-balencing automtic-recording potentiaters. The flow of heat through the outer surface vas calculated from measurements of the power dissipated in the electrical heating strips. This power was determined by measuring the resistance of ‘the strips and the current flowing through them, Thermocouples placed on both surfaces of the plastic base at a number of chordwise stations gave en indication of the heat flow into the model interior.