Icing on Aircraft Citation Snell, William E. and Hannon, Paul G. (1942) Icing on Aircraft. Other, California Institute of Technology. doi:10.7907/sdwr-7m10. https://resolver.caltech.edu/CaltechTHESIS:09262025-220053495 Abstract No abstract. Item Type: Thesis (Other) Subject Keywords: (Meteorology) Degree Grantor: California Institute of Technology Division: Engineering and Applied Science Major Option: Meteorology Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 3 February 1942 Additional Information: "Submitted in Partial Fulfillment of Requirements for the Certificate in Meteorology." Record Number: CaltechTHESIS:09262025-220053495 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09262025-220053495 DOI: 10.7907/sdwr-7m10 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17706 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 04 Oct 2025 11:43 Last Modified: 04 Oct 2025 11:57 Thesis Files PDF - Final Version See Usage Policy. 10MB Repository Staff Only: item control page

ICING . ON AIRCRAFT by Wm. E. Snell & Paul G. Hannon Submitted in Partial Fulfil~nent of Requirements for the Certificate in .Meteorology. Meteorology Dept. Calif. Inst. of Tech.· February 3, 1942 I. ETRODUCTIOK Icing on airoraf't becomes a factor of major importance in these days of all-weather flying. In . .certain oiroumstanoes aircraft may become heav- ily lofs.de,1 'v.rith ice, particularly, but not entirely. over leading edges of wings and other parts. and the effeot on flight may be serious. In mild cases it may only render the radio useless by ooatin~ the antenna and causing it to break avra.y. In other cases the carburetor intake may be chokeil. or the controls jammed by it, but the greatest danger probably lies in 1~he effect on. the performance of t he airora.rt. due to the sheer weight of accumulated ice. or mainly to the dist urbance of the a:erodynamioal propert_ies of the supporting surfaces ca.using loss of lift and setting up . vibrations sufficient in.magnitude to throw the aircraft out of control or even to cause structural failure. Vfui1e not all icing is as violent as this. mild oases can be very dangerous in that it may cause vital instruments to become useless. I THE PHYSICS OF ICE FORMATION In studying ioing, .there are three different types of ioe to be considered. 6ne is clear or glaze ioe which is olear and hard with a smooth and glassy appearance, although when mixed with sleet or snow it may have .a rough appearance. It is very difficult to dislodge when formed to any appreciable thickness. This type of ice is the :n1ost dangerous because of the above reason and also because of the blunt shape it assumes as it freezes on the leading edge of the wing or strut with its frontal area :enlarged. The second type of ice is called rime ·t oe. It is pure white, opaque. 2. a nd ,~r an ul a r in structure. I t c , nsists of tiny ice pel l ets vi\, :i_c '. 1 have li t tle cohesion or adlrnsion. Hirne builds up into s}w.rp n osed deposits · on leading edc; es and does not alter t Le form of tr,e airfoil to a ny extent, althouc;h its acldi tioc-: al -v.rci :fl . • t1nd t he increase in drag caused by it building up ve-r:y :nuch will be not i c,1 ci. Ord:inarily this type of ice does not build up to large amounts because it is easily shaken off by vibrationo For that reason if is seldor'.l a dangerous type of ice. However, at low temperature, the tenacity of this form of ice increases and the formation, if prolonged. soon ren.ch:;G C: n.ng:erous proportions 0 The third a_rid least d,;,n~erous type of ice is frost. It is a lir,h t; crystalline formation and never assumes any degree of magnitude. There are certain cases Nh ere this tvpe ·will be slie;htly dangerous because of 1 a light, rough deposit on an airfoil, destroyi r g some of its liftinr; characteristics. Tiovrever, the length of time that this depos:i.. t will remain is s h ort, because as soon as the airfoil reaches a temperature equilibrium with the surrounding atnosphere, the frost v,ill melt, providinr,: the temperature j_s above freezing. The t:rpes mentioned above a.re the pure types. In t h e case of the first two there can be, and are, many intermediate grades between the transparent ice and pure vrhi te ri:me. The most corrnnon type is of a semi-opaque ui.ture.., similar to partially melted and refrozen snow. Diagrams s h· • g the charac- teristic shapes of rime and clear ice will be found on th e following page. Most meteorologists agree that there are t-vm prerequisites for any type of serious icin g on aircraft. One is that there must be liquid water present and the other is that the temperature must be 34°F~ or below. There are special cases in vrhich these conditions do not hold, but the type of ice fonned is not dangerous to any lar 6 e extento Liquid water mus·t be present because preaipi tation in the frozen or solid Eorm before striking the aircraft will not adhere to it but i,rill be blown away. In order for this liquid water to be present, it must usually 3. ~ ~ ----~ P.C. . H . 4. be in the form o.f supercooled water. This is an i mportant fa.ct because, as it is known from experiments, agitation of this supercooled water will cause solidification. The necessary agitation in this case is accomplished by an 'airfoil .moving through this subcooled water and the resultant i mpact e>f the drops on the airfoil~ In determining the type of ice to be formed. the icing time. or the time it ta...lces for the ice to form is of 9rime importance. If the time is relatively large, clear or glaze ice will be formed, while conversely, if the time is small, rime will be the end product. .Bleeker has shown that the icing time is directly proportional to the radius of the droplet of subcooled water end that the proportionality factor involved is dependent ma.inly . Qn the vapor pressure existing i n the air surrounding the droplet. In order to demonstrate the processes involved in t he formation of ioe on aircraft, let it be assumed that there is water in liquid form, either ~s a cloud, or, perhaps, falling rain, at a temperature of -8°C. As soon as the plane strikes suoh a sub-cooled droplet, there is provided the necessary agitation and freezing begins. freeze a.t once, however. All of the droplet will not For illustrative purposes, which will not inval- idate the results in any wa-g, assume the droplet consists of one cubic centime,; ter of liquid water. It is known tha.t to fre e ze one c.c. of water • 80 calories must be reraoved therefrom. It is also knovm the.t at the instant the droplet begins to freeze. the temperature of the mixture of ice and water will innnediately rise to O0 c. This rise in temperature can only be accomplished as a result of heat being added, while at the same time heat is being ta.ken away from a portion of the droplet in order to bring about the freezing. Assuming that no heat is added from without. this added heat must be removed from that portion of the droplet which freezes. , 0 Thus the \ rise in temperature of the mixture from -8 c. will give a measure of the number of calories required to produce this heating and thereby indicate the fraction of the droplet which is immediately frozen on impact. In this 5. example, sinco ~-t requires 1 calorie -~o raise the temp erature of 1 c.c 0 I through 1 degr~e c., it will require 1 x 8 or 8 claories to bring the mixture up to o0 c. And, since it requires the c::-~penditure of 80 calories to I I : freeze 1 o.c~, ,cooling has been PFov;ided sufficient to freeze 8/80 c.o. or lOJ~ of the ,volume of the droplet. .•Thus it can be said that the a.mount , of ·water enc6untered oy the plane in the form· of cloud. particles or rain I j . .• . drops, and which is immediately f'rozen. is directly proportional to the degree of under..;cooling of the droplots. , .T his point.will be considered later ,· ~vi t h respect to tho size of the '.' ·,~op let · (~s mentioned\ earlier)• If it was only that portion of the droplet which ' :freezes on impact that ~o. run off as water, concerned ·us, and the rernain:i.ng portion was permitted the icing of aircra.f't would not be of much impprtance. • .rt is because a very larGe part of the rema.inine; i,vater also freezes that the i ,cing; coni ·11:; .is through the oper- ditions are considered suoh a hazard to flying. 1 ation of the process of evaporation that the re*ini:ng :,ortiop_ of each drop:i.et is largely turned to ioe. Consider again. the origino.l drop],et, 10% of iI which has beeri,. changed • into ioe. v;hile the remaining 90% exists as water i at a temperatur~ of o°Co It is here that the vapor pressure of .' the su?Toundine; ai~ is very i r~ortant. \ / Over this wat or at o0 c. t l\ere is a s,aturation i.vapor pressure of 6.11 mb., whereas the surroundinc cloud air i,vill have a ~. aturo.tion vapor pressure ! corresponding to the temperature pf -e 0 c.f or 0111y 3.12 mb. Thus evapor- 1 ation will take place fror.1 the water surf'.a.oe because of its higher vapor pressure as compo.rod · to the surrounding ,air. ' ' To aecomplish the evaporation I of 1 c.c. of' water it requires the aclditibn of approximately GOO calories of heat. To f reeze t he rem.ain;ng 90;t of the original drop let, only +xnOx. 9 wh ich equals 72 calories needs to be . remove.d ._ Hence it is 0Y1ly necessary · I I • to ove.p ora.te 72 GOO or 12Jo1 of tho orir;inal droplet. .iowever, a.s all heat 1 is added or subtracted from this dropl~ itse1:r.:1 sdme ef t~J .,W8.t,..-z- ,t1,' (ilLbe lost, and evec: less heat will be reqtri ;" ')d +,o' be taken awayo This runou:J;1ts 6. to about 2;~ so that the total amount of the drop to be evaporated is ap• proximately 10;~ . Under these conditions it is seen that, roughly 90% of the vre.ter encou..'11.tered by the plane would be frozen. nearly nine tenths of which is caused by evaporation., and this ~;ar·l:; that is forrn.ed through evaporation is the most dangerous clear or glaze ioe. Thus it is evident that the process of evaporation is of first importance in any consideration of icing conditions of aircraft. particularly because of the strong ventilation afforded by their rapid relative movement through the air. The size of the cloud particle or the sub-o·ooled water droplet has no infl~noe_ on the pa.rt of ea.oh that is frozen either on impact or through subsequent evaporation. As was stated before, the _p;y;.. tion of each droplet frozen on impa.ot is directly proportional to the d et}'i-30 of undercooling, and so it is true that the rate of evaporation of the remaining liquid portion,. and the total amount to be frozen, will be directly proportional to the degree of undcroooling. At this point it is well to begin considerinr; at -what degrees of suboooling certain size droplets can and ordinarily do exist·. The different types of ice have been observed to form most frequently from 34°:F'. to o0 Ii'., althour;h pilots have found ioe a.t higher and lovrer tempe'l4atures. But, it ·was not of sufficient amount to be hazSJ"dous. Aboi:tt 80;·; of the olear ice formation occurs above a temperature of 20°P. ·while about 55;j of the rime ooours below that temperature. Temperature is a minor consideration as oompared with other factors, though, in determining the type of ice, as long as i·b is vn.thin the 34°F. to o°F. range. Air turbulence is an important factor hCNrever, because larger drops will be o.ble to exist vri th increases in turbulence. One exp la.nation advanced for the fa.ct that most glaze occurs above 20°F. is that at lower temperatures large drops of vrater tend to freeze spontaneously. Tkerefore, only the smaller drops in which ,·-.tT!le forms a.re in a.blmdanoe. Below· o°F• the water 7. droplets usually beooi,1e so wiall beca:use of the rarity of v,at.er i n tho J.iquid forr:1 tr..at they are deflected with tl-:e air around the airfoil ' a nd other part s of tho plane-. 'l'he reason they are def l ected is beca use of the ir smallness t hey have not sufficient inertia to penetrate the airstream at enough of an anc;le to rn.ake contact with t i-:e plane. In going on from here vapor l,ressure and icinr.; time vrill have to be considered alo!lg ·with other factors. In theillustration civsm ~bove it ,ms seen that tbe difference in vn.por :,ressure vra.s an i,nport mt f actor !n the evaporation of the liquid ,vater and the l't'Ulu.ltant .f reezing; of the droplet. In this illustration the air vro.s assumed to have a relative humidity of lOOJ~• Observations ho.ve shovm that in such clouds vihere clear ice for1:1s.tions were reported, the relativehumidities ranr;ed f'rorn 51;{ to 1005s vn th the average being a.bout 90~ . ' Hith lower relative humidities, it can be seen that the vapor pressures will also be loYrnr so that there would exist a still i;reater diff'erenoe in vapor pressure between that for the water on the wing of the plane and that in the f':r•Je air. Hence evaporation and the resultant freezing would take pL.c •• ' )re rapidly. This explains the observed rapid formation of ice as a plane emerr;es from a 1,1 0ist oloud into the cor-1ps.ratively clear,. dry air above or·below. Under certain conditions such as that of heavy U..Ylderbooled rain~ aircraf't will fail to piok up large deposits of ice. radii and are underoooled only a If the drops have large ffJ\'1 degrees, e...11d the relative humidity is great. the ioini; time will be large. The ~_,art of the dro:.clets not frozen on impact have time to be blown away and if heavy enough rain is falling, the water '\-rlll even have ti.me -so collect a.".ld wash some of the aJ.lready . frozen ioe aWtl-Y• As can be seen, this will definitely not e;ive the air- foil a load of clear ice, as clear ioe depends upon many thin layers of water to be successively frozen over eaohother. At temperatures below· 20°F. rime i ec is most predominant because the small dropleta enable the icinr.: time- to become very small. almost ins tan- 8. taneous. Upon impact only a part of t he drop freezes, bu-!:; t here is so little water in the drop that it does not spread over a."11 appreciable area. Consequently the formation is of a. granular structure. The rate o.t which ice forHs on an airplane part varies greatly according to icing time, which varies vnth other factors as mentioned, arid with the speed of the plane through the. free air. Observations differ over ~~de ranges from less than 0.5 in. per hour to as l1igh as 2 in. per :minuteG pilots observations have indicated that a ra.t e of one in, per minute is not at all unusual. At the bee;inning; of this paper frost was mentioned as a type of ice that aircraft are troubled with at times. A few oommon causes for frost a.re: (1) a plane fillying through free air which is below the freezing point long enough to assume the temperature of the oold air, and then suddenly going into v1a17mer air so that the water vapor in t his warmer air is condensed on the plane as small ice crystals knavm as frost. This occurence is not at alldangerous because fl. S soon a s the plane comes to the same temperature as thest~rro1mding air, the frost will melt affa (2) If a. ple..no has been sto.ndinc in the open air lon;:: enoug}: to collect a deposit of fros~, it should be brushed etJf before take-off is attempted because t :1e collect :i on of frost produces a s urface that is 110t smo oth and destroys the aerodynamical properties of the airfoil t o sorn.e extent. Th is can become quite i1::portant in the case of heavily loaded pl anes t hat do no t have too much room i n v1h ich to take-offs (3) This third c a se is very simi.lar to the first case a:nd is usually o:rrly come across i n Vc1e high latitudeso If there is an inversion near the ground so that the tern;:i. is belO'w freozinc ., it is possible for an airplane to take-off and climb rapidly enough t o still b e very cold when it reaches t h e inversion and conmences to collect frost due to the lowering 0 of the temp. of t he warmer air to the dew point which is 32 F. or less~ If tho plane is e.t ma.xim1..ir~ rate of cli1nb, t1, is occurence r;1a.y cause the air- 9. f oil t o become i neffici cm l:; enough to cause t h e plane t o st all be ca,we the f orr~at i on i s qui te rapi d $ Summary Point s to rm:1,1··,°i) or about icing are: 0 ,l y to b e heavies t at t emperature, . , : rly c lose to t he (2) c :i.e,).:r ice r:,,ay b e expected in clouds with approprinte , '3mp e:ratures, wher e vertical c ur r ent s strong en ough to supp ort the largo:r cloud droplets, are found. i.e. instability clouds or turbulent airo (3) Clear ice predominates above 20°F. ( 4) Rime predominates below 20°F. ( 5) Rime may be· exp ect ed in clouds, with appropriate temperatures" 1i,r herein vertical currents are not strong enough to s upport t h e larger droplets. i.,e. stable type clouds. "P.G.H. 10. II THE .METEOROLOGY Q! ICE FORMATION If a pilot is familiar with the physical processes causing an encountered or forecasted icing condition, his plans for avoiding the region can be logically made. In this section of the paper, a number of meteorological situations with which ice is commonly associated are analyzed and the practical application of these studies to flight plans is suggested. In general, the following practical considerations should be realized for icing in addition to previously mentioned theoretical considerations: l. The danger of icing in regions above cloud forms is not 2. Ice may be formed in rain below a warm front provided great. the lower cold air has cooled the aircraft surfaces below 32°F. 3. Clear ice may be expected with appropriate temperatures and when sufficient vertical currents are realized from the following: (a) Convective action resulting from the heating of surface layers. (b) Convective overturning or the forcing aloft of warm air at the forward portion of the cold front. (c) Active upglide of warm moist air over a cold (d) The front. lifting of cold moist air under the warm front due to the convergence of streamlines along the warm front. 11. ( e) The lif ting of un stab l e a i r over mountain range s . (f) All of t he above conditions will be shot•m on the diagram as being within a layer of unstable air, or if in a stable layer, immediately above an unstable zone. .4. The diagram will indicate stable conditions within the zone of rime formation, wherein vertical currents are not strong enough to support the larger droplets. It should be noted here that in addition to the absolute amount of liquid water a cloud contains, the rate of deposition of ice depends upon the mass of liquid water traversed per minute by an airplane, and increases with the velocity of the plane. balloon, moving with the wind, rarely experiences icing. A free Captive balloons and kites exposed to prevailing winds undergo serious icing only when exposed to 'the falling of supercooled rain. Similarly, in forecasting the 32°F boundary from the ground, it is absolutely essential to remember that the boundary also varies with airplane velocity. Because of the impact and frictio n of air particles on the plane, other factors being equal, the temperature of the plane is always higher than the air temperature and the differential increases as the plane velocity increase s . Due to this fact, the 32°F level of a plane in flight is higher t han the meteorological 32°F level. In cloud flying at altitudes between 3,000 and 7,000 feet, the following figures are applicable: 100 150 200 250 . 300 Approx. lifting of .32°F boundary (feet) 350 800 1500 2350 3400 Flying speed (mph) During flight, the pilot should watch the free air te mperature 12.. closely and remember that it does not indicate the meteorological but the increased temperature due to plane velocity. This increased temperature rise in flight through a cloud is about 40}& lower than in clear air. Therefore a pilot, before flying into dangerous ice clouds, should seek an altitude where the thermometer reads 1°F to 5°F above 32°F (depending upon air speed). is in order to stay below the 11 veloci ty cloud freezing level 11 This as differentiated from the 11 veloci ty clear air freezing level. 11 As has been suggested previously. one of the most striking things shown by material assembled is that not one case of icing was recorded in which lifting by some method was not taking place. cases. however, were recorded in the Western States. These It has happened in other places tha.t a plane flying through a relatively warm moist cloud-without icing--has immediately picked up ice upon passing into relatively colder clear air due to the difference in vapor pressures in the warmer and colder air. This has been explained earlier in this paper and noted as not very important. Of 218 cases on which United Air Lines have definite information up to 1938, the following occurred: 2 in instability showers over flat territory. 34 in currents over mountains with no fronts nearby. 182 through or in close proximity to fronts (with orographic _ and instability effects added in most cases). The times when ice is formed from thick fog or from rain falling into colder air are extremely _rare in the Western U.S. region. Ice occurred only twice in air that was definitely known to be stable, and in a 13. few more t h.at may have been. Since about 85% of the observed icing of aircraft oc curs in frontal zones, it is evi dent that considera ble study mus t be given to the distribut ion of .ice forming propert ies in t hose zones. GOLD FRONTS Ice formation along a nd in advance of cold fronts incre ases as the instability and :nois ture content of t he air in tJ:,..e warm sector increases. Be cause of the unus1.1al ly steep slope of cold fronts~ the ho ri zontal extent of the i ce area is generally smaller than along warm fronts, but because of t he more rap id lif ti ng , vertical currents are usually greater and t he large r liquid droplets will cause severe ice formation. A safe fligh t pe.th t o avoid ·dangerous ice conditions is i llustrated in Figure -I .Where temoeratu.res are below 15°F, ice par• . ti cles will predominate i n the cloud_ mas s, 2.nd even if t he frontal cloud iness extends to higher al t i tud.es than s hmm, only light ice will be encountered unless conve ction is very st r ong. The advisab ility of flying wes t to avoid t he icing area~ adjacent to t he co l d front i s ev ident, for in climbing t o the ea st, the a re a of seve re ice fo rmation would be encountered. The !!18.ximum a ltitude would, of cour se , depend on the directio~ of fli ght and the top of the clouds. An icing situation t ha t is fr equent ly observed along the eastern slope of the Rocky Mountains occurs when the cold fr ont of a Pc a i r mass moves southwest and west agai nst higher gro u.nd . The example shown in Jrigure II is from actual reports of pi l ot s flying between Albuquerque and Kansas· Oity on morning of Decembe r 8, 1937 . 14. ------------rs-o 12000 - ---------21)0 PATH ~~=~~=------.3o" - - - - - - - - 32" .--c--._.- - - - - ~ - - - - 35"9 , T7 -::~ •(> .> - - \...,. -----\f'o .,,/ . , . - ~. - 'J.J)(X> - < - ~ -- - - · ';- ~ PT _- -4o'' ·; ' - . i l/0" NK. w.e.s. 14000 12.000 /CE ~ • 0 (0000 0 fLIG+IT 3{, " -PAl'H - - 8000 _,,.,,.- ( j L \ 6ooo <iJ:> ~150 r '- '- ) .-..... ~·, . .'--',,, ~---,_ ~ x. ,. -,,( / ·" ' ..._ v·C / \ \qo 4000 \. \ 0 1J>OO 110 AB ' 0 SCALE OT C J ') 200 /00 IN TC AC Hl'-loS AQ -.._, I F \<,URE II KA /"' \VY /'" AY F WD f osoocs-r Df.C. 8, 1'137 W-ILS. 15. •In this situa tion, a modif i ed Pc a i r ma ss i s lift;ed t o above its condensation level by the encroaching cold air, and in t hi s unstable s t ate, sustains adequate liquid particles at freezing temper·atures, The pilot of an eastbound flight reported that when he attempted t o let down from 8000 feet through the cloud mass at Amarillo, a 1/411 of ice formed in 50 seconds. Flight was then continued above the cloud mass until the cold front was intercepted at 10,500 feet where breaks in t!ie cloud deck were observed, an4 only light ice formed. The frontal clouds in this case were only 20 - 30 miles in extent. After the flight reached the fresh cold Pc sector, no ice was deposited, although snow continued to fall fro m the higher clouds along the frontal surface. ~ FRONTS In ]!,igure m is a cross section of a wal·m front as described by Byers in his 11 S:moptic and Aeronautical Meteor ology, 11 a nd is similar to an actual condition which occurred betueen Pittsburgh and Newark on December 11, 1933. The important features to be considered in this example are the temperature distribution, t he re gions of the cloud mass . where liquid particles and ice particles predominate, ancl the precipitation forms. The heavy bla ck line marks the limit of ice formation; its position in the cold sector is predica tecl on the freezing isotherm in the rain area, and in the warm sector includes all of the cloud layer belo,1 freezing up to where liquid c l oud. particles cease to exist I and only ice needles, flakes or dry snow preva il. Ordinarily, this change in form occurs between 15°F and 10°F unless strong convection prevails to above this level. In flying through a warm front of this type, it is evident 16. 14000 - -lo°- 1iooo- 0 10000- lcf- 8ooo-- zoo 6000- 4000- LJoo zooo- 5' 0 w 500 L.IM IT OF ICE 14ooo12.000- 10000- -1cf. --- ---- 0 10° .,,- eoooI / ...,.,,-------- LIC.UIO l' Ric OoM1r111r.-r E,ooo- 1o· / / 4,ooo- 4oo 2£)00- So" / w f\G-vRg IJl 17. tbat the only safe flight path is one that will avoid the area where liquid cloud particles predominate. This can be accomplished only by flying at high levels, at 10,000 - 12;000 feet, or in ice particles above 9,000 feet. At any level below 9,000 feet where supercooled liquid predominates severe ice will be encountered. In the case of a plane flying below a warm front, if freezing rain is encountered severe icing could be avoided by climbing~ if that portion of the inversion above freezing is intercepted. It is important to remember that only when near the surface front should the inversion be sought by climbing, or well in advance of the front where the region free of liquid particles can be reached without encountering the area where liquid droplets do predominate. WA.RM FRONT WITH OROGRAPHIC IUFLUENCES The warm front condition just considered may be transposed into mountainous country as illustrated by Figw.·e N. The most prominent difference is the loss of the lower section of the warm front sm•face. This brings the freezing isotherm and the area of the cloud mass in which liquid predominates closer to the surface. I.fore rapid convection, induced by the steeper slope of the ground, will tend to sustain larger cloud droplets and consequently increase the ice deposit per unit time. A higher mountain range results a.s shown in Figure .lJl • The important resulting change is the concentration of the icirig zone • against t):,..e windward side of the mountain and its extension across the top of the range. altitudes. Safe flight remains possible only at high 18. Widespread areas of icing conditions frequently exist in modified air masses over mountainous regions with no definite warm front, but in advance of a cold front. Figure V shows this con- dition as it would be indicated on the weather map, and in cross section. The air in advance of the cold front is NPp in which the moisture content has been increased through the evaporation of moisture from the earth's surface. Under such conditions, this air would have a conditionally unstable lapse rate, and once condensation starts, mild turbulence will hold the resulting liquid droplets in suspension. Thus in those levels above the freezing isotherm, a severe icing condition would result, as liquid water would be available in the cloud mass. As Figure "'Sl. shows, each range of mountains will act as a warm front, causing ice forming cloudiness to extend over the area, and with strong wind, ice formation would exist to the top of the clouds. The combined effects of cold front and orographic lifting of an air mass rapidly lowers the temperature of the ascending warm air mass and produces moder~te to severe ice, since the cooling and condensation process is intensified, and liquid cloud droplets will predominate in the cloud mass. OCCLUSION A cold front type occlusion existed in the Chicago area on the night of December 7, 1937, and the cross section (Figure "SZI) is based on data recorded by a pilot flying through the frontal surfaces. The most outstanding feature of this situation is the 19. ( < 1412.- 10- 8- j,.. ".,,.,_A."" ~ ... .-.Ir~ .. " ~ " '){}° KC BN CG- > ' Go AO C:Y l./30 CST !)J=C. 7 - l'J37 W.E.S. zo. low temperatures, below 10°F, in which icing occurred. As is to be expected, ice occurred only along the frontal surfaces, where the necessary conditions for formation are realized. Due to the low temperatures of the warm air mass, the amount of available water vapor is small, but it is evident that even with the low temperatures, liquid droplets predominated in the cloud mass. Flight through such frontal surfaces should be conducted in a manner that will reduce the time of passage to a minimum, that is, a rapid climb and descent through cloud masses. This is good prac- tice, providing the thickness of the cloud deck is lmown with a fair degree of accuracy. If the warm sector bad been solid cloud, heavy snow would ordinarily be reported at stations below that sector. In the absence of such reports, a relatively thin cloud stratum is to be expected. CONCLUSION From the analysis of these situations, certain precepts to be used in avoiding ice formation aloft are derived: 1. The regions of most severe icing are associated with frontal surfaces. 2. Vertical convection sustains large water droplets and thereby induces severe icing, providing temperatures are below freezing. Cloudiness associated with strong convection should therefore be avoided. 3. Temperature inversions to above freezing do not always exi s t above ice forming regions, even when freezing rain occurs. Escape from ice by climbing should not be attempted unless existence Zl. of above freedng temperatures at higher levels is definitely known. 4. I.::e evaporates very slowly in clear cold air and an ac- cumulation of ice induces rapid collection of more ice if descent is made through a cloud deck in which liquid droplets predominate and the temperature is 32°F or lower. 5. When uncertain of ice distribution in clouds, flight in clear air above the clouds is the safest flight path. The flight run on climb and descent should be through broken or thin cloud~ remote from the frontal surface. 6. A thorough study of the latest synoptic chart to deter- mine the position, slope and tendency of fronts before flight is attempted is the only means by which ice forming regions may be located . and the probable intensity of icing determined. 7. sleet, etc.) rhe falling of precipitation (especially, mist, snow, from a cloud is indicative of ice formation therein if the cloud is presumably or known to be at or bel~w 32°F. PRACTICAL SUMMARY I Sources of Icing 2!! Aircraft 1. Flight throUfJl subcooled cloud.a: This is the most common source of rime. 2. Free.zing ~ aloft Q!. .211 ground: A temperature inversion is usually found aloft with temperatures above 32°. This is most common source of glaze. 3. Freezing drizzle aloft Qr, .Q.n ground: Temperature inversion to ,bove 32° aloft~ dependable in this condition. 4. Mixture .2f freezing rain • .rn and occasionally sleet: be encountered either aloft or on ground. 5. Wet rn aloft Q!. .Q!! ground: This may Rough glaze results. Dry snow can also cause dangerous ice if the plane is removed from warm hanger. The initial melting. then refreezing of snowflakes may produce a rough surface. preventing plane from reaching safe flying speed on ~keoff .. 6. Carburetor icing: This may occur with free air temperature as high as 70° if heat is not applied. Considered important today with present type of carburetor in the larger and heavier aircraft. 7, Splashing of airport runwcty: puddle water .211 plane during takeoff when temperature~ below freezing: This may produce rough coat- ing on control surfaces which can seriously affect control of plane. 23 8. ~'rost formed .Q_!! win1-:;s '"nd control surfaces while plane .1§_ .Q.!! ground: This can produce serious difficulty on an early morning takeoff due to rough surface. 9. Frost aloft: This may form on a cold plane flying into a warmer moist stratum of air. This melts or evaporates quickly and is unimportant. 10. Propeller icing during runup ru!Q.. takeoff ill dense subcooled fog QB ground: Unimportant only so long as flight minimums restrict takeoffs in dense fog; otherwise power loss. 11. .:&"light through clouds 2£. rain with apparent temperatures 32° 12. 35°F: This is occasionally reported and is probably due to expansional cooling above leading edge of wing. But thermometers are often in positive error by as much as 5° due to frictional heating. II Weather Map~ Associated with Icing l. Cold front: The icing may be quite severe through a narrow zone if the warmer air mass is conditionally or convectively unstable and therefore producing a line of dense cumulus clouds. The worst arrangement is a cold front of this type which strikes the plane course broadside. If icing becomes severe in this case. the heaviest icing zone may generally be avoided by detouring slightly to the rear of the front. 2• Warm front in which the overrunning warmer air 1s u.nstable: The heaviest icing should be found immedia tely above the surface of discontinuity aloft; i.e .• in the upper part of the transitional air between the two air masses (in . stable air the , heaviest water 24. particles tend to settle to t he bottom). Severe icing is often escaped by changes of altitude of only 1,000 or 2,000 feet under these conditions as the clouds are often well stratified. In changing altitude the direction of slope of the frontal surface must be kept in mind, e.g., with a steep frontal slope produced by blocking of air on western side of the Sierra Nevada or Rocky Mountains, a pilot climbing eastward might climb at nearly the same rate at which the frontal surface sloped upwards. 3. ~ front_ i n ~ the overrunning warmer air is convectively unstable: In this case cumulus clouds will form on the frontal slope, as well as the stratus type, and moderate to heavy icing may extend up to 15,000 feet or higher. In general, the least icing and smoothest air is found beneath the frontal surface. 4. Upper fronts: Heavy icing may nearly always be avoided by flying underneath the frontal surface aloft and in the lower layer of stable air across the top of which the upper front is moving. 5. 11 ~ of Occlusion" fil: Eye of Low: This is often a zone of relatively heavy icing resulting from convergent lifting (surface winds converge on this area from all sides), and from frontal effects. If icing becomes too severe in this zone, the only • definite out or alternative is to turn back or to climb to a very high level.(?) Cumulus clouds ma~r build up to 15,000 feet or higher over a fairly wide zone at this time. Development is generally sud.den and without warning. 6. Orographic lifting: Some examples of this source of icing are: Wand NW winds carrying moist and unstable air of Pc origin up the west slopes of the Alleghenies; E winds carrying unstable 25. PA air up the east slopes of the Alleghenies in late winter and spring; and E winds blowing upslope from GI to CX and DV. Ac- cording to studies ma.de by J. A. Riley of the U.S. Weather Bureau. westerly gales over the Sierra Nevada in winter often cause a zone of severe icing to extend from Summit to 11,000 feet and to extend from 25 miles W of the mountains to 10 miles E of the mountains. At these times, ice is generall~ not a problem at 12,000 - 14,000 feet. Thickness of the icing zone is greatest with highest temperatures and highest wind velocities at Summit. 7. fu Mass Instability: Some examples are: Pc air which has had heat· and moisture added to the bottom layers in crossing the Great Lakes in winter; PA air with the same characteristics which moves in from the Atlantic Ocean in late winter and spring over t he LO - KY sector ( this air is often more unstable than the Pc type on the western slopes of the Alleghenies); any cold Pc air mass which moves southward over increasingly warmer ground surfaces. 8. Stagnant Map with weak occlusion~ obscure fronts: Charac- teristic of a once active low which has filled in. Freezing drizzle quite characteristic with this type and an inversion in temperature to above freezing aloft should rutl, be e~pected. Tops of clouds are generally 10,000 feet to 14,000 feet msl and definite layers are generally present. Zb . BIBLIOGRAPHY Periodicals 1. Analysis of the Problem of Ice on ·Airplanes, by W. C. Greer. Journ. of Aaron. Science, N. Y., Sept. 1939, V. 6, No. 11, p. 451-9. illus. 2. Ice and Icing. Bull. of A.M.S., Milton, Mass., Feb. 1939, v. 20, No. 2, p. 59-60. 3. The Application of the Harvard Radio Meteorgraph to a Study of Icing Conditions, by K. O. Lange. Journ. of Aeron. Science, Dec. 1938, V. 6, No. 2, p. 59-63. illus. 4, Ice on Aircraft: What the British Are Doing About It. of A.M.S., Milton, Mass., Nov. 1938, V. 19, p. 384. 5. Icing of Aircraft Antenna Wires, by Geo. L. Haller. Jour. of Aeron. Science, N. Y., Nov. 1938, V. 6, No. l, p. 27-8. illus. 6. A Study of Aircraft Icing, by H. L. Hallanger. Bull. of A.M.S., Milton, Mass., Nov. 1938, V. 19, p. 377-81. Tables. 7. The Triple Point of Water and the Icing of Airplanes, by D. Arenberg, Bull. of A. M.. S. , Mil ton, Mass., Nov.· 1938, V. 19, p. 383-4. 8. Ice Accretion on Aircraft (Notes for Pilots), by G. C. Simpson. Bull. of A.M.S., Oct. 1938, V. 19, p. 326-9. tables. 9. Some Remarks on Physical Aspects of Aircraft Icing Problem, by A. R. Stickley. Journ. of Aeron. Science, N. Y., Sept. 1938, V. 5, No. 11, p. 442-6. 10. Icing of Pitot Static Tubes, by V. E. Carbonara and A. G. Binnie. Journ. of Aeron. Science, N. Y., Aug. 1938, V. 5, No. 10, p. 400-03. diagr., illus. 11. Studies of Synoptic Free Air Conditions for Icing of Aircraft, by E. J. ?Unser. Bull. of A.M.S., March 1938, V. 19, p. 111-22. Bull. illus. 12. Icing of Aircraft, by E. J. Minser. Air Comm. Bull., Dec. 15, 1934, V. 6, p. 144-9. (Reprinted in Bull. of A.M.S., May 1935, p. 129-33. 13. Formation of Ice on Aircraft, by T. A. Aspell. N. Y., Dec. 1937, V. 31, No. 6. p. 23. Aero. Di 6es t, 14. Concerning the Ice Accretion of Ailerons. Interavia, Geneva. Nov. 6, 1937. No. 488, p. 1-14. Tables. (Abstract Journal of the R.A.S. • London, Dec. 1937, V. 41, No. 324. p. 1158) 15. Development of Deicer. Rubber Age, N. Y., Nov. 1937, V. 42, No. 2, p. 109-12. illus. 16. Thermal Ice Prevention. Np. 1503, P• 391. 17. Icing Measurements on Mt. Washington, by S. Pagliuca. Journ. of Aeron. Science, N. Y., Aug. 1937, V. 4, No. 10, p. 399-402. 18. North Atlantic Air Services; Precautions Agafost Icing, by J. H. Parkin. Engineering Jour., Montreal. Aug. 1937, V. 20, No. 8, p. 645-46. 19. Formation of Ice on Aircraft, by H. Noth and W. Polte. Journ. of R.A.s., London, July 1937, V. 41, No. 319, p. 595-608. Flight, London, Oct. 14, 1937, V. 32, Tables. 20. Aircraft Icing Zones on the Oakland-Cheyenne Airway; by J. A, Riley. M. W.R., Washington, Mar. 1937, V. 65, No. 3, p. 104-8. 21. Ice Pie Crusts, by Robert E. Horton. Bull. of A.M.S., Feb. 1937, V. 18, Uo. 2, p. 66-7. diagr., illus. 22. Some Notes on Icing. No. 1338, p. 37-8. 23. Ice Formation in the Atmosphere, by D. McNeal. Journ. of Aeron. Science, N. Y.. Jan. 1937, V. 4, No. 3, p. 117-23. 24. Carburetor Icing, by Alexander N. Troshkin. 1942, p. 160. 163, 252. Aeroplane, London, Jan. 13, 1937, V. 52, Aero. Digest, Jan. American ~ Byers, 'Horace Robert. Synoptic and Aeronautical Meteoroloff 11 The Formation of Ice on Aircraft." p. 227-237 (Chap. XIII. McGraw Hill, 1937. (Refers to ~Unser, E. J., Icing of Aircraft, Air Comm. Bull., V. 6, p. 144-9, Dec. 1934.) 26. Brunt, David. Ph.ysical ~ Dynamical Meteorology. edition. Cambridge, 1939. P• 54-6. Second 27. Petterssen, Sverre. Introduction to Meteorology. p. 94-7. McGraw Hill, 1941. "Ice Accretion," Z8 - 28. Taylor, George F. Aeronautical Meteorology. Pitman, 1940. p. 115-116, 131-3, 142, 150, 207, 313-180 29. Jordanoff, Assen. Through the Overcast. 11 Icing and how to Avoid It, 11 p. 92-99. Funk and \iagna.lls, 1938. 30. Circular N. 31. Washington, 1941. Sutcliffe, R. C. Meteorology for Aviators. 432 (A.P. 1699)). First American Edition. Company, Inc., 1940. p. 13~3. (Air Ministry M.O, Chemical Publishing Pamphlets 32. Adhesion of Ice in its Relation to De-icing of Airplanes, by A. M. Rothrock and R. F. Seldon. Washington, 1939. 12 p. diagr, tables. (N.A.C.A. Technical Notes No. 723)' 33. Ice Formation on Wings, by L. Ritz. Washington, 1939. 8 p. diagr., tables, (N.A.C.A. Technical Memorandwns No. 888) (From Lilienthal Gesellschaft fuer Luftfahrtforschung, Berlin, Oct. 1938, 8 p., diagr., illus) 34. Meteorological-physical Limitations of Icing in the Atmosphere, by w. Findeisen. Washington, 1939. 7 p. Tables. (N,A.C.A. Technical Memorandums No. 885) (From Lil. Gesell. fuar Luftfahrt,, Berlin, Oct, 1938) 35. Report on Ice ·Formation on Aircraft. Washington, 1939. Tables. (N,A,C.A. Technical Memorandums No. 919) 36, Effects of Ice Formation and its Prevention. 17 P• (N.A.C.A. 24th Annual Report) 20 P• Washington, 1938. Ice Formation on Aircraft and its Prevention, London, H.M. Stat. Off., 1938, 27 P• diagr. (Air Ministry Pamphlet No. 83) 38. Report of Ice Formation on Aircraft by the French Committee• for the Study of Ice Formation. Washington, 1938. 20 p. Tables. (N.A.C.A. rechnical Memorandums No. 919) 39. Report on the De-icing of Aeroplane Wings, Tail Surfaces and Control Surfaces. London, H.M. Stat. Off., 1938. 32 p. (Air Ministry Pamphlet No. 86) 40. Ice Accretion on Aircraft, by G. C. Simpson, London, H.M. Stat. Off., 1937. 14 p. (Meteorol, Off., Professional Notes, No. 82) 41. Ice Formation in Clouds in Great Britain, by W. P.. Bigg. London, H.M. Stat. Off,, 1937. 35 P• diagr. (Me,eorol. Off., Professional Notes, No. 81) 42. Contemporary Sta te of Ic1 : Frognosi s , by V. Nazarov. Moscow, Scient. Tech. Dept • . . o , 1938. 12 P• (C.A.H.I . Transactions, No. 1) 43. )istribution of Temperatures over Airplane Wing with Reference to Phenomena of Ice Formation, by Edward Brun. Washington, 1938. 31 P• illus., tables. (N.A~C.A. Technical Memorandums No. 883) (From Pub. Scient. Tech. Min. de l~Air, Paris, 1938, No. 119, 47 p.) 44. Experiments in Calculation on Air Current for Ice Forecasts, .by V. Nazarov. Moscow, Scient. Tech. Dept . . . • , 1938. 4 p. (C.A.H. I. r.rransactions, No. 4) 45. Ice Formation on Airplanes, by Clarence L. Solmson. Aviation, London, March 1940. 5 p. diagr., illus. 46. Icing of Aircraft in Weather Manual for Pilots (Technical Manual No. 1-230)~ War Dept., 1940. p. 221-36. 47. Ice Formation on Aircraft, in Meteorology for Pilots (Civil Aero. Bull., No. 25). Sept. 1940~ p. 124-29. 48. Fronts and Aircraft Icing, in Air Mass and Isentropic Analysis, by Jerome Namias. A.M.S., Oct. 1940. p. 118-21. (From E. J. Minser, in Bull,, Ma.rch 1938, p. 118-21). P• 128 (From E. J. Minser, BUll., May 1935, p. 132). Commercial