ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR THE THERMOMETRIC ICE WARNING INDICATOR W Wo JNDEISEN' (Das thermometrische Vereisungswarngerat) (Deutsche Luftfahrtforschung, Untersuchungen und Mitteilungen, No. 691)'.. TRANSLATED FROM CENTRA AIR DOCUMENTS OFFICE FILE ATI 55569 BYBo Ao UHLENDOIF MYRON TRIBUS Project M992B WRIGHT AIR DEVELOPMENT CENTER CONTRACT AF 18(600)-51, Es 0. No. 462-BR-1 August, 1952

.,~/

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN THE THERMOMETRIC ICE WARNING INDICATOR By W. Findeisen Contents:. Page I. Review of the Development 1 II. Equipment 3 a. Description of the Indicator and Principle of Its Operation 3 b. Sources of Error 7 c. Mounting of the Instrument on the Airplane. 8 d. Electrical Arrangement 8 III. Detailed Design 9 a. Transmitter 10 b. Indicator 12 I. REVIEW OF THE DEVELOPMENT During a conference called by the Lilienthal Society in May 1938 on the subject "Icing," it was brought out that there was an operational need for an ice warning indicator. There were several ice warning devices known at that time; however, these devices could not be used operationally. Almost all of them indicated the total amount of ice accretion.on the airplane parts up to any one time, but they failed to indicate whether the amount of ice was increasing or decreasing. Major Helm pointed out at the meeting that operationally it is of prime importance to have an ice warning indicator which permits one to determine the "icing condition," i.e., the time variation of the ice accretion. The indicator must, above all, show how quickly the ice accretion is occurring in the cloud layer through which the airplane is flying, so that the pilot can take the necessary measure for the prevention of further icing hazards (change in altitude, change in course, or monitoring of the equipment for the prevention of icing). Only such an apparatus deserves to be called a true warning device. Other.

N-R: IRIWNG RESEARC INSTITUTE UNIVERSITY OF MICHIGAN i diiors, whih: show the ice accretion only after the operation of the a irple da;been; impairedaree of small use in practice. i -se-e.:/ Mato. HrI elm' -s remarks:based on his great experience in practical f:lyi-flyng, uced me' in the years following to search for a method of measua:eet ich esan.n fulfill the stated requirements. First, I investigated wi iion obtained from' my researclh..on electrical characteristics.of:ic i- hih are related to the precipitation-static probleml could be;utili ld. obtain a serviceable ice warning indicator, uring the forma-:tion; f i-. on a airplane part, there are liberated measurable electrical.chargeshi:encha bew used for the activation of an ice.arning indicator.:ntl' nb o. I hv. renot been sucessful in establishing the above-mentioned.tigpei Qsprit.us in a simple:and serviceable form such as would3 be re-c quir4 for an indicator for practical useo ~Cf:t ii'D';- ence with systematic temperature measurements from aircraft,.u n-dUltertkni ne with0precision the difference between the icing:.l1y-: er^an. t i x-.degree isothenrm',? gavebime the idea in 1941 to measure the.temperateof icing.. If a thermometer bulb is exposed to icing, its tem-:peratulre wilbe increased considerably due to the liberated heat of fusion |of the water,.thus changing the temperature indication. The temperature inre c nues as long as icing continues. The amount of the temperat:urerise:.e dends upon the rate of ice accretion. By means of'measuremrents of theltemperature increase of a thermometer exposed to icing, we can de-:termin th. icing condition (the icing differential). It;. tt:was apparent that the temperature rise of the iced thermbmeter could be -copared with another thermometer whose measuring bulb was likewise exposed to the air stream but was not influenced by icing. This can be achieved by proper arrangement of the tetermometer bulbs' We must, however, consider the heating which occurs on the thermometer bulbs due to'compresion and. friction in the flowing air, for this heating is not the same on th differentlyexposed therometer bulbs, Through the diffrent.effects of ~pressure ad. friction on the tw thermometer bulbs here.an.our:;per-. ature:d.ifferences which are of -the same order of magitudeas thase which.an be p edu ce by different rates of icing. On the basis of..owledgeobtaied:Lo fr numerous flight i^nvestigations, carried out by the Cloud Investigailsti e in Prague we have succeeded ineliminatin this source -:of error. e also su.aeeded in doing way with other sources of errors -which: wl be discussed later We now have an ice warning indicator that I recordsI aimst without error the icing condition in units of "millimeters Iof- icper miile It further records the disappearance of ice by |W. iFid; isn, Met. Zeitsch.,, 1940, p. 201. -2ciC4.. (ut)-: 1209..- ur measurements are in complete agreement with those.1. 2'~~^.rl'~ 2 ~

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN sublimation or evaporation, for in this case.the two thermometer bu1lb indicate a temperature difference in the opposite direction. The device was developed at the Cloud Ihvestigation Institute of the Government Office for Meteorology (Air Force) in Prague, where there existed close coordination between laboratory work and practical flight testing. Government Consultant Walliier, and particularly Diploma Engineer Dr. Maass played an important part in this development. II. EQUIPMENT Just like an electrical airplane thermometer, the ice warning indicatOr co:sists of a transmitter, which is mounted on the outside of the airpla'e, and an indicator, which is mounted on the instrument panel. They are electrically wired to one another and connected with the airplane electricai system. (a) Description of the Transmitter and Principle of Its Operation The essential parts of the transmitter are two thermometer bulbs in the form of circular flat discs, 2cm in diametero The two thermometer bulbs are mounted, as shown in Figure 1, on a tubular cross in such a way that one is crosswise, and the other parallel to the direction of stream flow (forward-facing disc and side-facing disc). The two bulbs expose different cross sections to the stream and on entry into icing cond.itions, are affected in a different manner by the ice formation. On meeting icing conditions, an ice formation occurs on the leading edges of the wings, and simultaneously ice formations occur on the corresponding parts of the transmitter. Considerably more ice builds up on the forward-facing disc than on the sidewise-:facing disc. The thickness of the ice increases more quickly on the narrow side of the disc, but the fact that the icing surface for the forward-facing disc is greater is important for the determination of t t total amount of ice. As is well known during ice formation, the heat o fusion of the water is liberated with the result that considerable temperature increase occurs on the iced surface. The temperature of the iced thermometer bulbs 3Concerning the mechanism of ice formation, see L. Bitz, Jahrbuch 1938, Deutsche Luftfahrtforscchung, Erganzungsb.,j p. 180 3

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN are similarly increased. They take on an average temperature which is between the temperature of the surrounding air and the temperature of an iced surface, and the greater the surface exposed to icing, the more the bulb temperature varies from the air temperature. The forward-facing disc has therefore, a much higher temperature during icing than the side-facing disc. The temperature difference between the discs can be used as an indication for the occurrence of icing. Moreoever, this temperature difference is not only a warning signal but an indication of the rate of icing. The more ice is formed on a body the more its temperature must rise with respect to the surrounding air, since the temperature difference is brought about by the ice formations- liberating the heat of fusion. If the release of the heat of fusion ceases, icing also stops. The quantity of ice formed is proportional to the heat of fusion liberated and subsequently diffused, and this in turn is proportional to the temperature difference between the body and the air, as follows from the law of heat transfer. The temperature difference is therefore proportional to the rate of icing. As can easily be seen, the same difference in temperature which occurs between two differently iced bodies, also occurs in the case of the two thermometer discs of the indicator. The temperature difference is equal to the ice accretion intensity (with respect to time and surface), multiplied by a constant factor,. The icing rate is most appropriately given in units of millimeters of ice accretion per minute. The above-mentioned factor has a value of 1.3 mm of ice accretion per minute for a 10C temperature difference in the sketch arrangement. It is to be expected that the calibration factor is not strictly constant but has somewhat different values in different clouds, some of which contain preponderantly small drops, others chiefly large dropso This conclusion follows when one considers that the percentage of the drops which strike the surface depends upon their size and the cross section of the body.4 Accordingly, the relationship between ice masses on the two thermometer discs of the instrument must change, since they present differentsized cross sections to the air stream; likewise the measured temperature difference and the multiplying factor will change. Flight investigations have shown however that these variations are not important in a practical sense and can be neglected. With the help of the measurement obtained from the instrument in millimeters of ice per minute, one can estimate the icing on the wings, on the tail surfaces, and on the other parts of the airplane. It is a fact that the amounts of ice which form on these different profiles differ in Albreht Phys. Zeitsch(1931), 51 Ritz l.. 4. Albrecht, Phys. Zeitsch,, vol. 32 (1931), p. 5!l L. Ritz} l oc.

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN thickneess The thickest accretions occur dn those parts of the airplane which have the smallest cross section exposed to the stream. Consequently, the- ice accreftion rate is considerably greater than, for example, dh the wing surface; but it is approximately just aS larg as on the tail surfaces. The temperature difference between t te wo discs.of the transmitter remains constant as long as icing contiiues; i.e., as long as the ice is building up'o The ffect of the increas ing ice thickness is virtually negligible since at its exposed surface the same heat exchange occurs that existed at the very beginning of the icing of the AiScs. During- the ice buildup there continuously occurs on the surface of the ice mass a constant temperature of its outer surfaceo Consequently, the increased temperature difference occurs between the two discs in the same fashion as in the beginning of the icing process, when the discs were practically icefree. However, the temperature relation between the discs changes gradually with a heavy ice build-up and this change causes a diminution in the sensitivity of the instrument. When the instrument is covered with a thick ice layer, it yields approximately 50 per cent too low a measurement of the icing rate. This phenomenon is related to the geometrical relations between the ice masses and the discs and to the resulting heat-transfer properties of the ice-covered discs. The error decreases when in place of the discs of 2-cm diameter, larger and thicker discs are used, which affect the collection efficiency of the ice, even in heavy icing conditions, less adversely Since in the operational use of the ice warning device, it is less a question of the exact measurement of the icing rate than of the time variation of the icing process, and since only an approximate determination of the intensity is desired, one can almost ignore the diminution of the sensitivity at very high icing rates. Ordinarily, no flying takes place with very thick icing layers, and furthermore, through intelligent use of the ice warning apparatus, these thick ice accretions can be avoided. Therefore, it is not normally necessary to provide the transmitter with larger discs, which, naturally, when compared with smaller ones, have the disadvantage of a higher drag; this must be guarded against in the: actual use of this equipment. In the development of the apparatus, 3.5-cm discs were used in addition to those of 2-cm diametero Their use would be advantageous in all cases in which the icing rate, even after an icing of long duration, is no longer to be determined qualitatively but rather quantitatively. When icing ceases, the temperature difference between the thermometer discs disappears. This occur.s as rapidly as the thermal capacity of the ice-covered discs permits; practically anywhere between 5 and 10 seconds. The temperature difference, however, does not simply return to zero but -takes on a high negative value. This phenomenon depends primarily upon the changed meteorological conditions. Icing occurs normally in subcooled water clouds, and in these 5

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN clouds there exists a relative humidity of virtually 100 per pent. Above or near the clouds, the relative humidity is usually considerably-leas. There ice accretion not only ceases but the ice begins to evaporate. For this process, which results in a gradually decreasing quantity of ice, mch heat is required which results in a corresponding temperature e:asee on the layers The temperature disturbance is greater on the forward-facing disc than on the s.ide-facing dcis, so that the. transmitter reports an inverted. temperature difference and therefore a negative reading. Again the magnitude of the reading of the indicator is proportional to the change of the ice. Large negative readings mean a rapid decrease in the ice accretion. The diminution of ice is, furthermore, not only dependent upon the relative humidity of the air through which the airplane is flying, but also upon the flight velocity, since at high speeds in cloud-free air there is a considerable decrease in the relative humidity around the airplaneo5 For this reason, there is often a diminution in the ice accretion, and the instrument shows this through a negative reading Negative values occur even to a greater extent than with evaporation of the ice mass, if the ice melts as a result of an increase in the air temperature, for example, during descent. This process is easily understood, since in melting, heat is required. The heat requirement is in this case very large because melting occurs usually much more rapidly than epora-tion Negative values occur not only during the disappearance of previously present icemasses, but to a smaller degree at- timeseven fr- icefree transmitters. In the evaluation of the dependability of the indicator the following observation. is pertinent. these negative measurements occur only in flight through clouds where an ice accretion would melt, for example, in flight in water clouds above the freezing temperature Th impninging water drops do not freeze in this case and theref do not give up a heat of fusion. On the contrary they have a temperature which is lower than the surrounding airo The air which arrives at the disc at free-air speed is warmed through compression and friction due to the conversion of kinetic energy, as the air is intercepted. Warming of the-air is somewhat less during flight through clouds due to the action of the water vapor at the surface of the body ("wet adiabatic), but i:t is still considerably greater than the warming of the impiinging drops. The conversion of kinetic enrgy associated with the drops prodiuces only about one-half as great a heating since the specific heat of the water has a higher value than that of air, Under these circumstances a body which is exposed to the impingement of many drops of water must be colder than a body which is not so much wetted, -W. F indeisen, Forschungs-Uo ErfahrungSberichte des Reichswetterdienstes, Series A, No. 11, Berlin, 1941., \~ 6 ____________________________________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY. OF.MICHIGAN and the relative temperature diminution of the forward-facing disc of the transmitter can then be explained and the negative value understood. This phenomenon is more pronounced in thick clouds and at high velocity. The negative values remain after the cloud has been left behind, that is, as long as the thermal elements, especially the forward-facing disc, are still wet. The temperature difference is usually even greater than it was in flight through clouds. It disappears however much quicker than, for example, in the sublimation of the ice, and for this reason it i's unimportant for practical purposes. At any rate, the negative values indicate the absence of i ing danger. (b) ourees of Error In the consideration of the principles of operation of the transmitter, we must not overlook two sources of error, which may result in both positive and negative temperature differences between the thermometer discs Ieven-without icing, melting, or evaporation taking place at the transmitter. One of these two phenomena may be improperly indicated by the instrument. During the development and the technical perfection of the ice warning device, both these errors hadto be considered. I ~ One of these errors depends upon the thermal capacity of the thermal elements. If the thermal capacity of the two elements is not the ame, then during high rates of temperature change of the air which is flowing by (for example, during a dive), a temperature difference between the two discs may occur. As a matter of fact, the thermal time constants of the two discs are not equal as a result of their different placement in the air stream. The disc which is placed in parallel flow to the'ir stream is better ventilated and can therefore adjust to surrounding te'iesature changes more quickly. In order to eliminate this source of error, one tlight make this disc somewhat thickero Technically simpler is the methOd.of diminishing its effective surface through the addition of a heat-insua&ting material. In this way the time constants can be easily equalized, The other source of error results from an unequal warming of the thermemete r discs in the air stream. The thermal elements perpendicular to.the air streiI is heated to a greater extent through the compression.of the onoeming stream on its windward side than the dise parallel to the stre am is heated through' the friction of the air stream. As a result there appeas, for exape.,1 at o300 km/hr velocity, a temperature difference of pproximately O.5~C. The effect increases in proportion to the square of the flight velocity and must be taken into account at especially high velocities, This effeet'ts. celled, just as in the case of the-thrmal-capacity effect, through a partial covering of the surface of the disc with a heat-insulati.ng material. This is possible since the discs are exposed on their opposite sides to different amounts of warming through friction and compression~ For the forward-..~~~~~~

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN facing disc, there occurs a greater warming on the windward side than on the lee side. On the sidewise-facing disc, the upper surface, which is attached to the bracket, is warmed less than the lower surface. If the warmer side of a disc is covered, then the average temperature of the disc is decreased, since then the heat transfer on the cold side has a proportionally much greater influence. The covering of the discs can be done in such a way that both sources of error can be diminished at the same time. Especially effective is the covering of the upper surface of the parallel-placed disc. However, it becomes evident that for complete elimination of the velocity effect, the windward side of the forward-facing disc should also be partially covered. (c) Mounting of the Instrument on the Airplane The indicator must naturally be so located on the airplane that it is subject to the same icing as the leading edges of the wings and tail surfaces. It is immaterial whether it is mounted on a wing or on the fuselage, but care should be taken that it is not exposed to the exhaust from the engines. The thermometer discs should not be exposed to the direct rays of the sun, since significant erroneous readings can be obtained in spite of the very high air velocity around- the airplane. Though these errors do not occur during flight through clouds, in which we are primarily interested, they might diminish the confidence of the flight personnel in the ice warning indicator. For this reason the indicator should be covered with a radiation shield. The indicator should not be mounted too far back on the fuselage or wing surfaces. If the thermal elements extend into the boundary layer, in which, as is well known, there exists a decrease in velocity, then one can not expect error-free operatian of the indicator. It must also be noticed that, in the air which is flowing in the neighborhood' of the fuselage or the wing surfaces, there are considerably fewer drops tha poutside the boundary layer. For this reason, the support for the indicator should not be too short in relation to the boundary-layer thickness; otherwise the indicator will give too small a reading for the ice-accretion rate.'(d) Electrical Arrangement The temperature difference between the two thermal elements which are characteristic of any icing conditioui can be converted into electrical signals and transmitted to an indicator by known electrical techniques. - 8

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN For this purpose the temperature dependence of electrical resistance of the thermal elements is used. To obtain the maximum electrical effect, which can be recorded with sturdy electrical equipment, it is desirable to use as the temperature element a material with a larger coefficient of resistance. To this end semiconductors are much better than, for example, platinum; semiconductors, furthermore, can be used in the much simpler shapes, such as are necessary in the case of the thermal elements of the indicator, e.g., they do not require the winding of coils. To compare the electrical resistance of the two thermal elements one can use either opposing coils or a bridge circuit. The opposing coils have the drawback of high voltage requirements and must have a relatively high current flow through the thermometer elements which can upset the energy balance. The bridge circuits have a lower current requirement. The bridge circuit of the ice warning device is shown in Figure 2, The resistances of the bridge elements, R1 and R2, are of such magnitude that, when the thermal discs are at the same temperature, no bridge current flows. In this case, the mechanical and electrical zero point of the bridge galvanometer agree, and drifting of the zero point as a result of changes in the bridge circuit is eliminated. Through this arrangement, the instrument is insensitive to the voltage fluctuations which necessarily occur in flight. At the most, only moderate percentage-wise errors in the measurement of the ice accretion rate can be caused, which, however, are not normally important. The bridge circuit is adjusted through resistor R2. This also regulates the current through the thermal elements and therefore the sensitivity of the arrangement. The indicator for the bridge circuit is used to indicate the icing condition through proper calibration of the scale. Indication takes place in the manner discussed in section II-ao No difficulties of any sort are expected. It may be desirable, even with slow ice accretions, during which there is only a small deflection of the instrument, to have a clearly visible warning signal, For this purpose the bridge circuit is connected to a relay which lights a warning lamp whenever a small positive temperature difference occurs on the instrument. -.~ ~ ~ ~ ~ II. D.ESI III. DETAILED DESIGN The detailed design of the above-discussed'apparatus offered no difficulties. The two parts of the thermometric ice warning indicator, the transmitter and the indicator, can be built with relatively simple means. ______________________ 9:__________________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Th. e aim of the development was to make a device of a form suitable for flight use. The specific requirements were~ dependable operation, insensitivity to mechanical loads, small space need, small air resistance, easy mountability, and little use of current. Figure 3 shows the individual parts of the ice warning device the transmitter and radiation shield being disass:iebled Figure 4 shows the transmitter as it is mounted on the outside of the planeo The details are given below: (a) Transmitter The tubular cross shown schematically in Figure 1 is welded to a base plate. At the ends of the tubes are thethermometer discs, each mounted with one screw; hard-rubber discs serve as insulators. The tubes carry the electrical wires which lead to the thermometer discs. We used "Urdox"-resistors, produced by the firm Osram, in Berlin, as thermometer discs. They are made of the semiconductor magnesium-titanium nspinell"* pressed into a mechanically firm molding compound. They are shaped in the form of a ring.of 20 mm outer and 11 mm inner diameter, approximately 1o5 mm thick. On either side of the discs, wire is welded to form a circular terminal. The electrical resistance of the discs, as with all semiconductors, decreases logarithmically with increasing temperature. In the temperature region under consideration, the change amounts to approximately 300 ohms, and the temperature c fficient of the resistor is apprxim ly 4 percent per degree, which is somewhat more than 10 times th at.of plwt inul^m Experience has showwthat the discs can be facre with rather lar;e tolerances, i.eo, the resistance and tepertue oetf~ients vary considerably from sample to sample. For the ice aig dei how ever, it isnecessary that the two discs tf a transmitter hae t ftically the same temperature coefficients, since otherwise the normal temperature variations may cause anunbalance in the bridge (for example, fallowing altitude changes in flight), which could lead to erroneous deflections of the indicator gage. For each transmitter, therefore, pair of discs must be found th.at have comparable temperature coefficients. It is not necessary that the resistace of both discs be the same, sine an ineulity of resistance can easily be taken care of through an appropriate selection of the omensating resistors (R1 ad in Figure 2). Also, the teperture *Spinell is a type of alloy composed of a bivalent and a trivalent element and having a characteristic crystal structure —Translator's note, -;___________ 10 - -

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN coefficients of the pair of disce in different transmitters may be different, for this results only in a change in the.seitivity of t theransmitter, which can be compensated for witho1ut difficulty through proper adjustment of the calibrating resistor (Rv in Figure 2). As a matter of fact, due to the irregularit ies of the Urdox.-discs8 the- et of res.istors, R1, R2, and Rv, must be separately adjusted. for every pair of discs. The selection of the pair of discs is made.on the basis of a.caibration in a cold batha The calibration curves appear as straight lines on a logarithic plot and therefore can be eeasily established. The slope gives necessary information on the temperature coefficient for the selection of the d is cs In the transmitter the discs will carry between 2 and 4 milliampereso This.small current causes a heating of the discs less than o.l~Co The measurement is not upset through this heating since the two discs are heated by the same amount. The potential at the discs is of the order of one volt. In general, it is somewhat higher than the electrolytic polarization voltage, and therefore care should be taken that no electrolytic connections are made to the surfaces of the discs, which -are often covered with a water film, for in this way erroneous readings may be obtained. The discs were therefore covered with an insulating layer of lacquer; Konstantol 706"T made by the fir of Tebas, in Prague, served this purpos. The lacquer dries at 150~C and gives a good protective coating, The adjustment of the.discs to equalize the thermal time constants and the heting ing the air stream was obtained through a partial covering with hard- rubber rings' These were inserted on the appropriate sides of the discs: namely, on the windward side of the forward-facing disc and on the top side of the s idewlse.-facing disc between the mounting screws (or mounting tubes) and the discs. Since the thickness of the discs varies, a particular size of cover cannot be used with all the transmitters to give the.same..adjustment.o The errors which occur as a result are quickly determ d through tests in a fast air stream, and consequently the covers can be adjus ted as required. As s shown n i the figures, the transmitter is provided with a radiation shield, which is mounted on the base plate of the tubular cross. In this way, the tubular cross and the discs are protected against great mechanical loads, as occur, for example, during flight operations. The indicator is mounted on the outside of the airplane along with the radiation shield. On all airplanes which do not have a motor located on the fuseage, it is advantageous to mount the indicator either below or on the side of the forward part of the fuselage because there the mounting can be carried out with least difficulty and the wires are shorter. ~ - ~~11.~

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN A three-strand cable connects the transmitter -with the indi.cator. In a new variation of the transmitter the resistance coil of the bridge are mounted on the transmitter, namely in a smallcapsule which is located on the mounting plate of the tubular cross an extends into the outer ak in of the airplane. With this arrangement a two-strand cable leads from the transmitter to the indicator and. a second cable to the airplane electrical circuit. (b) Indicator For an indicator no special device is required. We may use a moving coil galvanometer, which has been developed for other purposes on a production scale and is mechanically suitable for use on an airplane instrument panel, without additional adaptation. The instrument must be contained in a standard housing. For the sake of space savings the small standard housing of 57 mm diameter was used. Figures 5 and 4 show such a unit made by the firm of Hartmann and Braun, in Frankfurt am Main, called type "Hukka " It requires 0.37 milliamperes for full-scale deflection, and its resistance is 140 ohms. A s:imilar unit of the same range and resistance is said to be manufactured by the firm of: Philips, in Prague. The divisioning and marking of the scale of the indicator is adjusted to the particular apparatus. Its purpose is to show the icing conditions at a glance. As.hown in Figure 5, the indicator has its zero point approximately oe-third from the left of the scale. It maintains this position as long as the flight is through clou-free air. During fight in rainor in water clouds without dangerof icing, the indicator deflects to the left. These deflections, however, are relatively small and do not reach the left end of the scale.. They indicate, as shown on the scale, "No icing." The indicator moves sharply towards the left when an ice accretion which has already built up on the airplane disappears in flight through ice-free layers, whether inside or outside a cloud, due to melting or vaporation. The indicator then moves to its left most position, and the scale therefore states "No icing. Ice disappearing."; The principal part of the scale to the right, which, to indicate danger, is inscribed in red, shows the icing rate in millimeters of ice accretion per minute. The sensitivity of the circuit is so adjusted that the deflection to the end of the scale at the right corresponds to an icing rate of 6 mm/min. As mentioned above (;Section II a), this numerical value is correct only for thin profiles. On the parts of the wing near the fuselage thec iclisn-a.e.rably cosdei:ess A l-/min deflection indicates 12

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN;very rapid icing. It can occur only in hu.ge towering clouds wheref however, occasionally the scale is overshot. At todayT- flight speeds icing rates are normally approxiaately 2 to 3 am/min. Depend ing on the icing sens itivity of the type of airplane, there is more or less time to change altitude or course to find a way out of the icing layer.6 The transition from a layer with icing danger to a layer without icing danger is very clearly seen on the icing indicator, since the pointer does not only return to its zero poslition but rather goes to the lefto If new icing occrs, the indicator again goes to the corresponding position on the right. The indicator device shown in Figures 3 and 4 has on its rearward part a capsule in which the resistors of the bridge circuits are located (see Figure 2), These resistor are small spools of approximately one centimeter wound with coils of constantan wire. The two bridge resistors R1 and R2 have the same value of resistance as the thermometer elements, i.e- approximately 300 ohms. The calibrating resistor Rv has approximately 3000 ohms. Since the resistance must be adjusted to: the peculiarities of the pair of thermal discs, each set of resistors goes only with one pair of discs. Putting the resistance elements in the cockpit indicator has theref ore the disadvantage that the indicator and the transmitter cannot be interchanged. This disadvantage is eliminated in the new arrangement of the ice warning device in which the set of resistors is located at the transmitter. The housing of the indicator is consequently somewhat smaller and it requires only a two-strand.cable which.connects it with the transmitter. In the design considered first, two two-strand cables are necessary, one going to the indicator and the other to the airplane electrical supply. See in this connection, D. (tLuft) 1209, "Vereisung^*.. — 13 5

AIRPLANE FUSELAGE OR WING SURFACE THERMOMETER ELEMENT m-oo THERMOMETER ELEMENT ICE ACCRETION Figure 1. Schematic presentation of the transmitter of the ice warning device. Two disclike thermometer elements are exposed to the air stream in such a way that on entering icing conditions more ice will form on one element than on the other, RV FORWARD- FACING SIDEWISE-FACING DISC DISC + 24 VOLTINSTRUMENT BOARD BATTERY Figure 2. Bridge c ircuit of the ice warning device.

I 77 F igure 3. P ice warning WIN Figure 4. Transmitter with indication shield. -50=0

GR^TION IN t 2 3 4 NO ICING i ICE DISAPPEARS Figure 5. Scale of the indicator. Inscription on left is black, on right (region of danger) in red.

3 9015 02111 31625 3 9015 02826 3625