REPORT No. 126 AERONAUTIC INSTRUMENTS SECTION IT ALTITUDE INSTRUMENTS x Four Panta AERONAUTIC INSTRUMENTS SECTION BUREAU OF STANDARDSREPORT No. 126. ALTITUDE INSTRUMENTS. PART I. ALTIMETERS AND BAROGRAPHS. By A.B. Means, H, B, Haxmicrsox, snd W. @, Buowmaczer, INTRODUCTION. This report is Section II of e series of reports on aeronautic instruments (Technical Re- “ ports Nos. 125 to 182, inclusive) prepared by the seronantic instruments section of the Bureau of Standards under research authorizations formulated and recommended by the subcommittee on aerodynamies and approved by the National Advisory Committee for Aeronautics. Much of the material contained in this report was made available through the cooperation of the ‘War and Nevy Departments. ‘This part discusses briefly barometric altitude determinations, and describes in detail the principal types of altimeters and barographs used in aeronautics during the recent war. This is followed by a discussion of performance requirements for such instruments and an account of the methods of testing developed by the Bureau of Standards. ‘Tho paper concludes with 1 brief account of the results of recent investigations. Altimeters and barographs ere used for determining the altitude of aircraft ebove the ground, the former giving direct indications on a dial, and the letter furnishing a continuous record on a chert. For accurate measurements of altitude, reference must also be made to thermometer readings of atmospheric température, since the altitude is not fixed by atmospheric pressure alone. ‘This matter is discussed in the following section on barometric altitude determination. Obviously, any instrument which purports to determine altitude solely from observations of atmospheric condition is liable to some uncertainty unless very complete observations of the te of the atmosphere aro taken throughout tho entire period of time and over the entire region of space corresponding to the flight, including simultaneous observations on the ground. Thus the determination of. the most probable eltitude from barometric observations requires an elaborate system of observations and intricate computations. Fortunately for many purposes extreme accuracy is not necessary. It is possible to make altitude observations by direct reading of the altimeter without any computations or supplementary observations, pro- vided an accuracy of 10 or 15 per cent is deemed sufficient. This is done by having the alti- moter dial graduated directly in altitude units, based on the assumption of some particular constant value for atmospheric temperature. BAROMETRIC ALTITUDE DETERMINATION. It is comparatively simplo matter to deduco mathematically the relation connecting pressure, temperature, and altitude for a perfectly stationary atmosphere treated as an ideal ges. If proper account is taken of the actual variation of temperature vertically throughout the eir column such a relation will be sufficient for most purposes of altitude determination, This relation may be referred to as the general altitude equation, which may be written in either of the forms Bf’ ap i fs a) aE hog, @ E tog, @ or so1502 REPORT NATIONAL ADVISORY OOMMGTTER FOR AERONAUTICS. Byidontly the choice between the two forms in any practical problem depends whether the absolute temperature, @, has been given as a function of pressure p or as a function of altitudo A. In the foregoing equations 2 is the ges constant which occurs in the characteristic equation of an ideal gas— a= po= Ro @ where v is the specific volume (reciprocal of density); g denotes as usual tho acceleration of gravity; and p, represents the barometric pressure at the ground, where the altitude J is zoro. An example of e case where the temperature is given as a function of pressuro is found in Radau’s law. This is the empirical law assumed in graduating the dials of most of tho altime- ters manufactured in France. Evidently when @ is given explicitly as some algebraic function of p with numerical constents, the integration can be completely worked out. Upon sub- stituting suitable numerical values for the actual gas constant of the atmosphere & and for gravity g in appropriate units, there results « definite working formula connecting altitude with pressure. ‘Radau’s law makes the temperature a linear function of pressure. ‘The advantages, how- over, of assuming the temperature to be a linear function of altitude rather than pressure have recently been urged by some authorities, and in this case the other form of the general equation would be used. ‘The chief value of the general relation given above is, howover, for the determination of probable altitude in cases where the temperature variation hes not been given mathematically but by actual observations taken throughout the flight. In such instances the integration may be worked out graphically or by some numerical step-by-step method. Unless the temperature has been actually observed all the way up the air column, it is of course impossible to arrive at any very accurate eltitude determination. So mauch for the use of the general altitude equation es it stands. ‘This relation may also be simplified or extended. An exemplo of tho altitude pressure relation when reduced to its very simplest form is found in the graduation of British and American altimeters. In both cases this simplification consists in assuming for the eir column e strictly uniform temperature of 60° F, (10° C.). ‘Thus the temperature comes outside of the integral. sign, making the altitude proportional to the logarithm of the pressure ratio. ‘The value of the ges constant and other constents adopted from the Smithsonian Meteorological Tables as a busis for American practice are such as to lead to the formula: = 62000 logs 722° @ In this equation, i represents the altitude in feet corresponding to a pressure p in inches of mercury. All American altimeter dials are graduated in accordance with this formule. British altimeters aro graduated by means of substantially the same formule, the difference being negligible for practical pufposes. Woile British and American eltimeter scales are in substantial agreement, those adopted on the continent of Europe differ considerably and are not interchangeable, ‘When the available data and the importance of the determination warrant an oxtension of the general altitude relation, some of the factors which may well bo teken account of are the following: Variation of barometric pressure on the ground during the time of flight; of the gas constant or even departure from the law of an ideal ges, due to humidity; effect of the wind in modifying the normal static distribution of pressure; small correction for variation, of gravity with eltitude, eto. ‘The foregoing discussion is intended to indicate some of the difficulties inherent in the determination of altitude from a knowledge of barometric pressure even if the prossure- measuring instrument itself were mechanically perfect.AUIITUDE INSTRUMENTS. 508 PRINCIPLES OF ALTIMETER CONSTRUCTION. The altimeter is not necessarily different from any other form of aneroid barometer except that the principal seele on the dial is graduated in altitude units, while the pressure scale is usually omitted. As a matter of fact the altitude scale is usually also e scale of equal parts. ‘This is a great convenience because it would otherwise not be methematically legitimate for the altitude scale to be rotated in order to make the zero setting at the stert of the flight. This error in design occurs in the older forms of aneroid barometer, such as the familiar pocket size used in surveying and mountain climbing, in which the pressure scale is usuelly equally spaced. Consequently the 1,000-foot intervals on the eltitude scale are about twice as close together in the neighborhood of the 20,000-foot point as they are at sea level, in accordance with the logerithmetio relation (equation 4). Suppose, now, that the pressure on the ground has shifted half an inch from its normal value. With the older style aneroid, having a movable altitude scele on the bezel ring, the zero setting would naturally be made by turning the scale half en inch of pressure, which means about 500 feet at sea level but about 1,000 feat at the other end if the total altitude is, say, 20,000 feet. Thus an error of 500 feet is inadvertently introduced. For accurate work an unequally spaced altitude scale should not be rotated. It should be clamped in position and the altitude found by subtracting the initial eltitude reading (on the ground) from the final altitude reading (at the top of the climb). ‘This awkward procedure is eliminated in the altimeter, where the equally spaced divisions of the altitude woale do permit rotation. The action of an altimeter is otherwise similar to that of any aneroid barometer. . ‘The essential parts of an aneroid are the pressure measuring element, the transmission or multiplying mechanism, end the indicating element. In addition, auxiliary devices aro some- times introduced to compensate for possible sources of error, notably temperature changes. ‘Throughout the discussion of the different altimeter designs which follows later, the description will in every case be token up successively with reference to these four features; namely, the pressure element, the mechanism, the indiceting element, and the compensation. ‘While one of the earliest forms of aneroid employed 4 Bourdon tube for the pressure ele- ment, the forms adopted in aviation invariably have for the pressure element some combina tion of flexible diaphragm and stiff steel spring. The springs serve to weigh the total force due to air pressure acting on the diaphragms, after the feshion of any spring balance. The larger the diaphragm, the greater will be the total force available to actuate tho spring. As a general principle, in instruments of this class a lerge fore action is desirable in order that friction and similar sources of error may have a relatively small influence. ‘The diaphragm boxes or capsules are commonly called vacuum boxes, because there is usuelly a fairly high ‘vacuum inside the box. This would not be necessary so far es the mechanical operation of the instrument is concemed, provided the temperature could be held constant. ‘The practice of exhausting the boxes to a high degres is merely to avoid change of pressure duo to expansion or contraction of the confined air when heated or cooled. The diaphragms are corrugated in order to make them more flexible. ‘They are commonly constructed of German silver (nickel brass) although other materials have been used to a less extent. ‘The mechanism for multiplying the motion of the spring, so that a large deflection of the pointer may be realized, hes been developed in « great variety of forms. ‘The numerical multi- plying power ranges in different types from about 200 to 800. Various combinations of levers, sometimes together with cam motions or gearing, will be found in the descriptions which follow. ‘The indicating element in eltimeters consists ordinarily of a pointer moving over e gradu- ated dial. Various methods are used for graduating the dial and controlling the zero adjustment. ‘The microscopes, verniers, and micrometer screws which have been used on surveying encroids are not found in the aviation type. Optical indicating devices have been tried but are not in common use. : ‘Tho compensation of altimeters almost invariably includes statio balancing of tho mechanis by countereights so as to overcome as far as possible the error due to inelination504 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. of the instrument, or to linear accelerations. If not otherwise specified it may be assumed that such belencing hes been satisfactorily accomplished. Angular accelerations aro also always presont at the instrument board of an airplane. Compensation for such accelerations is more difficult, and seems not to have been soriously attempted except in @ recent French instrument. Some of tho eltimeters developed at the beginning of the war were rejected on account of faulty balancing in this respect. ‘Temperature compensation is frequently provided for by the bimetallic method, or by rogulat- ing the amount of airin the vacuum box when the instrumentis manufectured, or by both methods. Tho bimetallic method consists in providing a compound bar in some part of the lever system; for example, a brass lever with a thin steel strip welded onto it. Change of temperature pro- duces curvature in such a bar if normally straight; and it should be so designed that, the amount of curvature per unit change of temperature will compensate for the deflection of tho pointer which would otherwise take place at-constant pressure due to temperature change. Tt will be observed that this form of compensation does not necessarily compensate for the chango in the sensitivity of tho instrument with temperature. Even if an instrument is so compensated that the pointer will not respond to temperature changes while the pressure is constant, it is still possible that the amount of deflection for unit chango of pressure will be different at differ ent temperatures. Without bimetallic compensation there may be a partial degree of compon- sation, satisfactory over a vory limited temperature interval at atmospheric pressure, pro- ‘vided for by the admission of air into the vacuum box before sealing it up. ALTIMETERS OF THE D-SPRING TYPE. SHORT & MASON, One of the oldest and best known aneroid movements is tho Short & Mason (fg. 1). ‘The Tycos altimeter, tho Neko produced during the war, and soveral other British instruments such as tho A.'T. Reynolds, and S. Smith & Sons altimetors aro substantially identical in general design with the Short & Mason instrument 60 far as a brief description would show. ‘These instruments form the principal group coming under what Follucltuanimass, i known as the D-spring type. ‘The Short & Mason instrument will bo taken as an example and described in detail. Other instruments need be deseribed only so far as they deviate from this one. igure 2 shows a top and side view assembly drawing of the Short & Mason altimeter. ‘The large stool mainspring (10) hes somewhat the form of a letter D in cross section and is known to the trade as a D-spring; hence the designation of this type of construction. Historically the original form of this mechanism, in its mein features, was devised by Naudet in Franco. ‘The present Naudot altimoter will be deseribed later. "The pressure element in the Short & Mason altimeter consists of the D-spring (10), to- gether with one vacuum box which is coupled to it by means of a knife-edge. The vacuum box is made up of two corrugated diaphragms which sre formed so as to overlap at tho rim where they are joined together by soldering. ‘The bottom diaphragm is attached to the baso plate (6) by moans of a very stout screw end nut (15). If the top and bottom were not forcibly held spart during exhaustion the vacuum box would collapse under atmospheric pros- sure. ‘Tho tansion in the mainspring is so adjusted by the carriage screws (8 and 9) as to hold the top and bottom diaphragms approximately parellel under normal atmospheric pressure— if anything, slightly concave. If the air pressure is now decreased, as it will be in flying to a higher altitude, the atmospheric pressure on the surface of the vacuum box diminishes, re- Iieving the tension on the spring and allowing the free end to deflect upward. ‘This movement is of the order of 1 millimeter in 20,000 feet of altitude. ‘The movement of the free end of the mainspring is transmitted to the pointer by means of the multiplying mechanism. This, in the Short & Mason instrument, begins with the mainALTITUDE INSTRUMENTS. 505 lover (92) and a floating link (20) which connects the end of the main lover with the regulator spindle (17). The link is not attached to the rigid part of the regulator spindle but to the regulator spring (18), which is a flexible etrip offset from tho axis of rotation of the regulator spindle by a distanco fixed by the position of a amall adjusting screw shown in the drawing. ‘Thus the upward motion of the D-spring due to a diminution of air pressure causes the floating ink (20) to pull up on the rogulator spring, thus rotating the regulator spindle counterelock~ ‘wise in the drawing. From the middle of the regulator spindle a straight lever (19) projects in an upward direction. A fine brass chain (48) transmits the movement of the upper end of this lever (19) over to the chain arbor (49). Tho chain winds end unwinds on this arbor, which is connected with e hnirspring (48) for taking up lost motion. ‘The chain arbor forms solid part of the pointer spindle (47). ‘Thus the motion of the vacuum box and spring is trans- mitted successively through tho main lever (32), the link (20), the regulator spindle (17), the lever (19), chain (48), and chain arbor (49) to the pointer or hend (52) which moves over the graduated dial (4). ‘From the side-view drawing in figure 2 it will be seen that the initial position of the lever (19) hese good deal to do with the uniformity of the altitude scale. ‘The multiplying power may be constant in an ordinary aneroid, but must be made to vary loge. rithmaticelly in the eltimeter. This is not always a simple matter and constitutes one of the reasons for tho great variety of multiplying mechanisms in ex- istence. ‘The indicating element of the Short & Mason instrument consists of the pointer (62); the dial (4), ‘which is movable and which carries the altitude greduations; and the dial (6) which is stationary but notgraduated. Teeth ere cut in the outside cireum- ference of the movable dial which mesh with pinion (1). ‘This pinion is actuated by means of the Knurled knob (28). In order, therefore, to adjust the zero point of the altitude scale to coincide with. the position of the pointer at the stert of a flight it is only necessary to turn the knob (28). ‘The temperature compensation of the Short & Mason instrument is accomplished at sea-level pres- sure by the usual bimetallic method. The main lover (82) has a thin steel strip (38) welded onto the top side. For this reason tho mein lover is commonly known as the compound ber. ‘The details referred to in figure 2 may also be recognized in the photographic view of the dismantled Short & Mason altimeter (fig. 3). ScuEDER. In the Schneider altimeter the prassure clement is substantially the samo as in the Short & Mason type. In place of the Maxwell point, groove, and plane mounting of the spring cerriage, the carrisge is held in position by three screws, two of which end in points, and the third in polished plane. Lock nuts are provided on the carriage screws. ‘This company has also made aneroids with steel vacuum boxes. The multiplying mechanism, althoigh similar in arrangement to the Short & Mason, is different in one respect: The chain arbor block is machined with a spiral groove of variable radius 50 as to produce e uniformly spaced altitude scale.506 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. ‘The indicating element is similar to the usual form but without the stationary dial. ‘The temperatare compensation is bimetallic. Ordinarily tho bimetallic bar hes been incor- porated in the main lever, but in some samples the lever from the regulator spindlo has been used instead. ‘Tho Naudet instrument, (6g. 4) is simply a modification for aeronautie uso of the original aneroid barometer of the D-spring type, which they were tho first to develop. ‘The deserip- tions of the Short & Mason instrument ebovo will serve to givo an iden of the essentials of the ‘Naudet construction. Tho pressure clement differs from the Short & Mason in that tho main spring earrings is swung on eylindrical bearings. ‘The modified form of the regulator spindle in this instrument should be noted (fg. 4). The regulator spring (18) is free to move about an axis perpe dicular to the axia of the regulator spindle (17) and is therefore capable of adjustment in two planes. ‘This adjustment is obteined by means of screws (21) and (218). Only one dial is provided, which is movable and is graduated in meters. The usual bimetallic com- pensation is provided. GC A ad » e he o Feet & Mars eso disarm fo. t--Dlara can of Nant atte covompes, Another French altimeter is vety similar to the Naudet. In addition to the movable alti- tude scale, there is also provided a stationary dial graduated in millimeters of mercury. The pointer of this instrument may be set to the true pressure by means of the adjustment. of the carriage screw. This is accomplished by adjusting the regulator screw through « hole in the bottom of the instrument case, aa is the practice in surveying and weather aneroids. NUDSEN: ‘This instrument of Danish construction has a pressure element similar to the others except that the center portion of the spring is machined out. The-multiplying mechanism is also different. In place of the usual regulator spring thero ise pin which slides in end ont of a transverse hole in the regulator spindle and which is clamped in position by a set screw.ALTITUDE INSTRUMENTS. 507 ‘The movable eltitude dial graduated in meters is rotated by means of the knurled bezel. It moves over « stationary dial, a small sector of which is graduated in millimeters. ‘The tem- perature compensation is bimetallic. . OERMAN ALTDUETER, ‘This altimeter (monogram GL) has been found in two slightly different forms (figs. 5 and 6). ‘The carriage screws point upward and fit in conical cups in the carrjage. The carriage-adjusting serew is placed in the base of the instrument and acts against the arm of the carr . ‘The chain and hairspring are errenged contrary to the usuel manner so that the pointer moves clockwise for increasing altitude. ‘The main lever connecting the D-spring with the mechanism is threaded, thus allowing edjustment of its length. ‘The regulator spindle end spring are made in one piece of metal. ‘The hairspring is of phosphor bronze. ‘emanate. Fie. 6-—German linet with dal remot In the two forms referred to, only the indicating element is essentially different. One type has both a stationary and movable diel. The movable dial is rotated by pushing on a projecting knob which comes out sideways through a slot in the case. In a Inter form there is only a single dial, which is fixed to the case. ‘The entire aneroid mechanism is revolved by fa simple form of gearing show in the’ photograph (fig. 6) operated by « knob on the outside of the case, The usual bimetallic compensation is provided. orto noms. ‘The pressure element of this altimeter has carriage screws pointing upward and fitting in conical cup and slot. The carriage adjusting screw is in the base of the instrument as in the other German altimeter. There is a link between the vacuum box and the D-spring as in the De Giglio altimeter. ‘The regulator spindle and spring are made in one piece of metal. The indicating element is provided with both a pressure scalo graduated in millimeters and a movable altitude seale graduated in meters rotated by gearing and knurled knob. The temperature compensation is bimetallic. HELICAL SPRING TYPE. BICHARD AND SDMA INSTROMENTS. ‘The helical-spring type is well exemplified by the Richard instrument. It differs from ‘the D-spring type in that the mainspring consists of a vertical helical spring under compression. ‘The spring is not directly coupled to the vacuum box, but the downward pull of the vacuum bos is transmitted to the helical spring through a flat plate having its fulerum on two steel points. ‘The movement is shown in figure 7. The disphragm (11) is coupled to the fiat plate lever (7) by the knife-edge (14) working on a steel collar (14a). The tension of the helical ‘spring (10) is varied by the screw fitting on which the spring rests. This can be adjusted by ‘means of a screw reached through a hole in the base of the instrument. ‘Any movement of the diaphragm (11) is multiplied in the ratio of about five to one at the end of the long lever arm. ‘This movement is trensmitted through the multiplying system through the pin (20). ‘This pin strikes the curved surface of a cam causing it to rotete about the axis (25).- Attached to the cam is a flat brass pin (19) whose motion is transmitted by508 EEFORT NATIONAL ADVISORY COMMITTEE YOR AERONAUTICS. contact with an arm projecting from the sector gear (48). ‘The teeth of this sector gear mesh with « smell pinion on the pointer spindle, Backlash is taken up by a spirel hairspring. ‘Thus the movement of the top of the vacuum box is transmitted to the cam through contact with 2 pin; the rotation of the cam and upright pin (19) transmits the movement to the sector goar, thence to the pointer spindle. ‘Tho indicating elemont consists of a movable dial graduated in meters revolving over a stationary dial graduated in millimeters. ‘Temperature compensation is accomplished by the relative expansion of the brass end steol parts of the movement, but without any bimetallic bar. waror. In this instrament, also of French construction, special care is taken with the mounting of the fiat plato lever or carriage (sce fig. 8). The carriage screws aro of stool with sharp points and are held in small cylindrical brass posts inserted into the base plate. ‘The bearing surface for the carriage screws are small eylindricel pieces of steel inserted into thé carriage plate and are cupped to receive the points of the car riage screws. The connection of the di- aphragm to the car rage plate should be noted. There is a square hardened steel pin inserted trans- versely in the dia- phragm pillar similar to the ordinary knife- edge. The surfaco of this pin facing the di- ‘phragm is cupped to receive two steel points which are in- e7-Dapaataicartatiete, Sertedin thecarriage. sy. 5 sein eae inae 1a. 1-—Digas Bihar Tenants, rms std held distended by « helical spring as in the case of the Richard. ‘The tension of this spring can be adjusted by means of a scrow which raises and lowers a brass fitting on which the spring rests. Referring to figure 8, there is shown a screw in the offset from the circular head of the carriage plate. Tn the end of this screw steel pin is inserted eocentzically. ‘This stecl pin rests on the polished face of the multiplying lever. By rotating the screw that holds the steel pin the multiplying ratio can be adjusted. ‘The multiplication of the motion of the carriage is accomplished in the following manner: The steel pin in the carriage rests on the polished steel surface of a lever shaped somewhat like a-bell crank. ‘Tho other arm of the bell crank is in the form of a stecl knife-odgo, and takes the place of the upright pin of the Richard instrument. This lever is counterbalanced. ‘The steel knife-edgo is in contact with the tail of a toothed sector. The motion is transmitted to the indicating hand by the seotor meshing with a pinion mounted on the same arbor shaft with the hand. A tension is maintained on{this lever by moans of « light steel spring. Baoklash is taken up by a steel hairspring.ALTITUDE INSTRUMENTS. 509 The indicating mechanism is a fixed dial graduated with o pressure scale in millimeters. Over this fits a movable illuminated dial provided with a crown gear. ‘The dial is rotated by means of e small pinion mounted in tho instrument case which is turned by a knurled thumb screw. No bimetallic compensation is used in this instrument, ‘The pressure element of this instrument differs from the Richard by providing en extra connecting link between the vacuum box end the fitting which carries the knife-edge. The mechanism is similar to that of the D-spring altimeters except thet the rotation of the hand is clockwise for increesing altitudes. . ‘The indicating element has both a stationary pressure scale graduated in millimeters and movable altitude scale graduated in meters. ‘Thero is no bimetellic temperature compensa- tion, but the statio balancing is dono as in the Richard, by the in- sertion of a strip of lead into tho back end of the flat plate lever. wmoreexica. ‘The Filoteenica altimeter, also an Italian make, is very similar to the De Giglio except thet the movable diel is rotated by turning the knurled bezel instead of a geared thumb screw. ‘OTHER TYPES. ‘The pressure clement of this Franch altimeter (fig. 9) consists of vacuum box with an intemal spring. This construction is “8 familiar in barographs, but unusual with altimeters. ‘The multiplying mechaniam sterts with a long lever consisting of a brass rod into which is inserted transversely a long steel up- right. Offset from this lever is an arm (19) which engages the top 7 of the vacuum box. ‘The bearings (26) of this long lever are mounted. in a bridge over the vacuum box. The long steel upright actuates a geared brass sector (48). This sector meshes with a pinion on 4, the index arbor. Tho zero of the instrument is adjusted by tuning a steel screw which raises or lowers one end of the mechanism bridge. ‘C] Backlash is taken care of by a small epiral spring. ‘The indicating element has been made in two forms, in each 2 ut of which there is both a stationary and movable dial. Tn oneform *- TAPS ast tinte the movable dial giving the altitude scalo in moters is rotated by gearing actuated by a knurled thumb screw. In the other form the movable dial is revolved in a knurled bezel. seemer. ‘The pressure clement of this movement is made up of a battery of two vacuum boxes with internal springs which are mounted on a U-shaped brass frame. ‘This U-shaped casting holds the entire mechanism and is screwed to the case of the instrument. A bimetallic bridge spans the two legs of the U-shaped piece and supports the multiplying mechanism in the center. This mechanism consists of a carefully machined helical groove in the pointer arbor. This arbor is froe to turn in a polished hole in the bridge. It is caused to rotate in proper relation to the movament of the upper vacuum box by the vertical movement of e pin traveling in the helical groove. ‘There is a spiral hairspring to take up backlash. ‘Tho pressure element of this altimeter consists of a single vacuum box with an intemel spring. Any motion of the upper surface of the diaphragm is communicated to the multiplying mechanism by means of the jeweled bearing soldered to the top of the upper diaphragm.510 REPORT NATIONAL ADVISORY COMBUTTRE FOR AERONAUTICS. ‘The design of this instrument is ynique and offers several advantages in the, adjustment for calibration. ‘The mation of.the diaphragm actuates a short’ lever arm projecting from tho regulator spindle, which rests on the, jeweled bearing of the diaphragm. . This lever is punclied integral with the regulator spindle, and, sinco the material is of spring stecl, considerablo adjust~ ‘ment of its length can bo accomplished by the lateral motion produced by a rogulator sezew. Tho vertical movement of the diaphragm serves to rotate the regulator spindle, which carries with it an upright. The movement of, this up- right in tur rotates a calm to which it is con- nected by alink, ‘The cam rotates about a ver- tical exis. By moens of a hairspring mounted on the cam shaft lost motion is taken up and a slight pressure maintained against the jeweled bearing. In contact with the carn surfaco there is a small roller boaring which is attacliod to the tail of a geered sector. ‘Thus any rotation of the,cam causes a eprresponding movement of the geared sector. ‘The cams so designed that with increasing altitudes there is a gradually increas- ing movement of the sector for equal movements of the diaphragm. ‘The geared soctor meshios with the pinion on pointer spindle, 2 ‘The stationary and movable dials aro constructed in’ the usuel way and the latter is (urned by a knurled knob. ‘ra nGumas atte wits doable ston. (GERMAN ALTIMETERS WITH! DOUBLE ACTION. ‘The pressure cloment of this altimeter consists of two batteries of two diaphragms each. ‘The springs are internal. The vacuum boxes are mounted edgewise 60 that the direction of their motion an deflection is.in a plane parallel to the dial of the instrument. “The two Datteries deflect in opposite directions as shown, by figures 10 and 11. : —- ‘The deflections of the diaphragms cause a rotation of the multiplying lever (17) about an axis, perpendicular to plane of motion of tho diaphragms, the two arms of this lever being connected by links to the diaphragms. ‘The motion of the multiplying lever (17) is transmitted by moans of a short pin (21) to.a second slotted lever (20) carrying the sector (48), The pin (21) slides in this slot when the lever (17) is displaced. ‘The pin (21) is attached to lever (17) by a lock nut and washer and is adjustable in the slot at the end of lever (17), thereby making it possible to change the multiplying ratio of the ever system and hence the defection of the pointer. ‘The geared sector engages with a small pinion on the pointer s spindle. A hairepring is mounted .on the indicating I arbpr. a ‘The indicating clement hes an unequally divided fe-1!—Digas met ct Cea ante th altitude scale which is fixed in position. The entire 7 inside mechanism is rotated by means of a goared sector engeging in a pinion mounted in an offset in the instrument case. The eltitude scale is considerably cramped at the high altitude end. Tho sotation of the hand for a change of altitude of 1,000 feet near the 25,000-foot point is only one-third thet at sea level. ‘Temperature compensation is effected by the general design of the lever system.ALITTODE. INSTRUMENTS. 51 PRINCIPLES OF BAROGRAPH CONSTRUCTION. Barographs may be described with reference to the pressure clement, multiplying meche- nism, recording element, and compensation. ‘The pressure element is invariably such es to give a larger amount of deflection then in the case of eltimeters. This is necessary in order to reduce the demand on the multiplying mechanism, es will be seen. ‘The increased sensitivity of the pressure element is usually attained by having a battery of vacuum boxes one on top of another. ‘The steel springs may either be internal or external in arrangement, the former being more common. In barogrephs the mechanism has considerably less multiplying power than in altimeters. ‘This is necessary in order to insuro a sufficiently firm and powerful movement of the trecing point; otherwise the unavoidable friction of this point on the chart might cause trouble. Various optical and other devices have been tried for the recording element, some of them with complete success, but none of these has como into very common use. Ordinarily the recording element consists of a pen-and-ink record on paper chart. The chart is placed on & revolving drum whose time of revolution ranges in different instruments from a half hour to 24 hours or longer, according to the requirements of the work. The scale provided by the chart is rarely as open in the barograph as in altimeters. ‘The term altigraph has frequently been used for consistency with the term altimeter to designate a barograph whose chart is equally spaced with respect to altitude, and some charts are provided carrying only the altitude scale. ‘Temperature compensation is usually accomplished in barographs only so far as is possible by means of admitting air to the vacuum boxes. Bimetellic compensation is rerely provided. Balancing is also seldom attempted, for it is expected that the barogreph will be held in an upright position, INTERNAL SPRING BAROGRAPHS. THE RICHARD BAROGRAPH AND SISILAR INSTRUMENTS. ‘The most widely used and probely one of the oldest types of barographs is the Richerd. ‘The Green barograph is very similar to the Richard except for the uso of a tubular-sheped pen instead of the usual form consisting of a V-shaped trough. ‘The prenrare clement la a battery of fwo internal spring dlaphragms attached securely at the baso to a flexible flat steel plato or spring. A square-headed screw operated by a key adjusis the height of the boxes above the base plate by causing an up-and-down motion of the spring end serves as a means of adjusting the zero setting of tho pen. ‘The Richard move- ment is shown in figure 12. ‘The motion of the upper diaphragms, pillar (A), is transmitted by means of a link (B) to & multiplying lever which oscilletes about the exis (0). ‘The long pen arm is actuated by means of a link connecting it to the multiplying lever and an upright (D) from the pen arm. A flat spring is provided which keops a slight tension on the pen erm and takes up backlash. ‘The pen arm of the recording element is long flexible fat spring steel shaft. This long thin erm is in turn festened to the last rigid lever arm by two serews. It is sprung away from this lever arm and then brought back toward it by « thumb serew. ‘This method of regulating the tension in the pen arm serves to adjust the amount of pressure at the contact of the pen with the chart. ‘The pen cen be removed from the surface"of the chart by means of an arrestment actuated by a shaft projecting outside of the ease. abe, ‘The pressure element of this barograph is similer to that of the Richard except for the zero adjustment being made by # knurled thumbscrew. ‘The mechanism serves to transmit the deflection of the diaphragms to the pen arm in the following manner: The movement is first transmitted to # multiplying lever by mesns of an upright extending from the uppermost vacuum box. Both arms of the multiplying lever are512 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS, adjustable and are held in place by set screws. The pen is actuated by means of a link which ‘connects this multiplying lever with an adjustable arm attached to the pen arm. Thus a further degree of adjustment is possible than in the case of the Richard berograph. The recording clement is similer to the Richard except in certain details. ‘The pen arrestment is operated by a lever extending outside the case of the instrument, to which is attached a brake serving to stop the clockwork whenever the pen is thrown off the chart. The mechanism and base plate slide into a wooden case. The pen erm is counterbalanced. A barograph developed in England and made for a limited timo in this country is shown in figure 18. The pres- sure element consists of a battery of threo internal spring vacuum boxes. Unusual attention was given to the quality of the clockwork and to care in construction. ‘The multiplying lever (C) is mounted on an arbor having conical beatings sus- pended on o bridge between two up- rights, . The height of this mounting can be varied by turning a knurled thumbscrew (8), thus adjusting the zero of the instrument. ‘The multiplying lever is connected by meens of a link (D) to the pen arm. A helical spring ‘keops the tension on the pen arm and serves to take up backlash. ‘This instrument was the first to provide a true altigraph scale. It has also an unusually open scale. This was mado possible by # very high drum. The period of rotation of the drum in some ‘forms is as short as 30 minutes, adepting it especially for’ performance testing and experimental work, In this barograph a bime- tallic bar is provided in the multiplying lever (0)." It is the only barograph known in which bimetallic compensa- tion has been ettempted. ‘The pressure element of this barogreph is similar tothe Richard, One end of the multiply- ing lever is attached to the vyaucum boxes by means of steel link, The other end is attached by an adjustable pin to the arbor of the pen arm. AcouNT. 2s iio daca osm atadh Cover romored, ‘Tho pressure element consists of » battery of two internal spring vacuum boxes mounted rigidly on the base. ‘The uppermost vacuum box is connected to the multiplying lever by means of a steel link, ‘Tho multiplying lever is a long brass arm, the axis of which is held in bearings mounted on & bridge. ‘The height of the bridge ebove the base is adjustable by means of a thumbscrew, thusALIVTUDE INSTRUMENTS, 518 providing for the zero setting. ‘The thumbscrew is placed in a yoke mounted on two standards. ‘The adjustable bridge is prevented from getting out of alignment by two guide pins placed through the yoke. ‘The multiplying lever is counterweighted and actuates the pen arm by means of a stecl link. srorraNt ‘This Swiss barograph hes a battery of two internal spring vacuum boxes mounted on the base plate by means of a short upright post. ‘The uppermost vacuum box is connected to the multiplying mechanism by @ link. ‘The multiplying lever is elaborate in construction, offering facilities for ready adjustment. ‘There is « short arm made up of en adjustable pin threaded on the end. This pin is kept in tension. by & helical spring, ond it is held in a cylindrical-shaped fitting bored so as to receive the pin. The mounting for the multiplying lever is similar to that of the Short & Mason barograph. ‘The pen atm is attached to the multiplying lever by means of a link. ‘The multiplying lever is counterwoighted. NUDsEN. ‘This instrument of Danish construction has for its pressure element a battery of three ‘vacuum boxes with intemal springs. ‘The zero adjustment is made as in the French barographs. ‘Tho movement of the vacuum boxes is transmitted by @ multiplying lever, one end of which is attached to the uppermost-vecuum box. The other end of the multiplying lever is atteched to the pen arm by means of a con- necting link. The length of the arm of the multiplying lever can be edjusted by means of a slide which is held in place by a set serew. ‘The length of the lever on the pen arm may be adjusted by means of a set of holes for the link pin. ‘Tho pressure eloment of this barograph (fig. 14) is composed of a sylphon mult capsule instead of the usual battery of sep- arate vacuum boxes. The syiphon consists in effect of seven capsules in series, elthough made of one piece of metal. Instead of the usual clliptic springs there is a single intemel helical spring. Tn laboratory tests this instrument hes shown exceptionally small elastic hysteresis and after-effect errors. "Tho top of the sylphon is attached to the multiplying mechanism by means of a link. ‘This link is attached to an arm projecting downward at an angle from the main spindle. ‘The pen arm is mounted on this seme spindle. ‘The zero adjustment is accomplished by « thumb- serew adjustment on the pen arm which permits a rotation about the main spindle. 1a. 4B apn are ‘This instrament, of German construction, has « battery of four internal spring vacuum boxes mounted on « flexible base ao that their height may be adjusted by turning a imurled ‘thumbscrew. ‘The deflection of the vacuum boxes is transmitted direotly to the pen arm by means of a connection from the uppermost vacuum box to an adjustable sheft through the pen-arm arbor. ‘The pen arrestment is operated by a small knob which extends out through the door of the wooden instrument case. ‘The clockwork is wound by turning the drum in the opposite direc- tion from that in which it is driven. 20167—23—88514 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. EXTERNAL SPRING BAROGRAPHS. RICHARD POCKET BAROGRAPE. ‘This instrument is very much smaller then the ordinary barographs. ‘The pressure clement consists of a battery of two vacuum boxes without intemel springs. They are prevented from collapsing by e holical spring somo distance ayay, which is mounted in a vertical position and which acts upward against a at, plate which serves as.one of the multiplying lovers. ‘This part of the construction is somewhat simi- arto the helical spring eltimoters. ‘Thore is an adjustment for tension of the helical spring by means of a square shank scrow operated by # key. This servos es a zero adjustment for the instrument, (Soo accompanying photograph, fig. 15.) Inserted into a projection from the multi- plying arm or fat plate, above referred to, there is a pin which points downward. This pin acts con tho curved surfaco of e cam. Any movo- mont of this cam causes a spindlo to rotate which earties an upright pieeo of wire serving as alover. This wire rod makes a sliding con- tact on a fin which is mounted upon tho pen arm. ‘The baek-lash of tho pon arm is taken up by & hairspring. ‘The recording element contains a mechu- nism for periodically raising and lowering the penin relation to thechart. ‘The chart is not in contact with a single cylindrical drum es in the larger berographs, but is stretched betwoen tivo small rollers so that the surface receiving the reoord is flat. The rollers are forced apart by a spring which can bo released in order to remove the paper. The pen is raised from the chart by a slotted bar in which it is free to slide. This slotted bar is raised and lowered by clockwork; thus the record consists of a scries of dots, the pen being of the paper the majority of the time. ‘This device enables the pen arm to take up the most accurate position without restraint due to friction at all times. The necessary firmness of contact formaking » legible record is secured momentarily by the periodic action of the slotted bar. ‘Fs 15 Ricard poke baoragh. ‘The pressure messuring element of thisin- strument (Gg. 16) is very similar to that of the Richard pocket berograph. It consists of ‘vacuum box of the extemal spring type which is held from collapsing by « helicl spring ~~ ye.1—mupuaoehaimareapieaiews, placed at the lower end of o T-shaped plate. me cud aena This plato serves, asin the pocket barogreph, as one of the multiplying levers. Any movement of the diaphragm of this altigraph is multiplied further by means of another multiplying lever (B) and the pen arm (C) which are connected together, and the multiplying lever to the T-shaped plato by means of small links. ‘The construction of the multiplying lever makes-possible con-ALEITUDE INSTRUMENTS. 515 siderable edjustment of the multiplying ratio. It is yeally & regulator spindle, and the ratio of the length of its two arms can be varied by turning a knurled nut with the fingers. The pen and pen shaft are similar to that used in the Richard berograph. Except for a slight modification of the clockwork, this instrument is similar to the Richard. PERFORMANCE REQUIREMENTS FOR ALTIMETERS AND BAROGRAPES. ‘The performance requirements of altimeters and barograpbs are determined, on the one hand, by the degree of accuracy necessary for the particular use contemplated and, on the other hand, by the sources of error existing in the instruments under conditions corresponding to those which will be experienced in fight. ‘The conditions occurring in flight which may cause errors are extremes of temperature, inclination, acceleration and vibration of the instrument, and rapid change of pressure tending to cause a time lag in reading. ‘These conditions must be reproduced in the laboratory and the instruments tested more or less completely under each condition, depending on the accuracy required and whether or not the instrument is of a new type of construction, ‘Schedules of permissible errors for various conditions and altitude ranges have been edopted in making tests at the Bureau of Standards. ‘Those will not be gone into numerically here, but the testing procedure will be explained and the performance characteristics to be observed in the test will be definitely stated.516 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. METHODS OF TESTING. ‘Th apparatus employed for the pressure and temperature tests is essentially the same for all kinds of tests on aneroids of the aviation type,'and is shown in the three accompanying photographs. Figure 17 shows various types of containers found useful in the calibration of aneroids. ‘The bell jar to the right of figure 17 is used for experimenting on aneroids. It is equipped with electtical connections so that a smell fan motor, heater, otc., may be put inside; also it is pro- vided with small copper tubes for carbon dioxide refrigeration, so that extreme low tempera- tures may be maintained. Other apparatus not shown in the photograph has been used for routine tests of ancroid barometers at temperatures as low as around —40° C., especially for the flight history tests. ‘The container in the center of figure 17 is convenient for routine testing of smell groups of ancroids and especially for checking the readjustment of aneroids following repair. It may be covered either with the adjacent flat glass disk or with a small bell jar, as shown in tho rear of the photograph. ‘The container at the left of figure 17 is one of & type designed for testing a largo number of altimeters at the same time. It consists of a large bell jar, inverted, and rotated in a motor- driven support. It can be read either through the top or the side. ‘The vacuum connection is made through a special fitting in the knob of the jar. ‘The next photograph (Gg. 18) shows a temperature chamber in which any of the foregoing containers may be placed or two of the inverted bell jar type. By the use of the brine systemALIYTODE INSTRUMENTS, 517 and heaters, any temperature down to —10° C. may be mainteitied, and slightly lower temperi- tures may be seoured by using calcium chloride to remove the frost from the brine coils. The instruments are read through a double glass door. The containers are rotated by a motor that is controlled by the two Imife switches shown in the lower foreground. Figure 19 shows the mercurial standards end a vacuum control board of special design used in aneroid testing. Below is « large reserve vacuum tank. ‘The tank is of value in mini- ‘a 19—afaeria standard vacua ctrl oar. mizing the effect of slight leaks that may occur in the system. ‘This arrangement of barometers ‘and vacuum control board and tank makes it possible to produce different pressures in several different containers simultaneously, so that instruments of different ranges may be tested in the shortest time. Also it facilitates the calibration of the mercurial standards, which were submitted to the Bureau of Standards during the war in considerable numbers. ‘Tho vibration apparatus as shown in figure 20 is of special design in that the motor is not attached to the vibrating frame. This lengthens the life of the motor, and therefore the effi-518 REPORT NATIONAL ADVISORY COMMIFTEE FOR AERONAUTICS. — * cienoy of the vibrator. The amplitude, frequency, and direction of the vibration can be varied at will, meking it possible to subject the instruments to any desired vibration. ‘The chain-testing device in figure 21 is shown to illustrate some of the special problems which may arise in sample instruments. Several different styles of chains were tested on this apparatus for a comparison of their performance and dural re Dating bord, Special apparatus wes also developed to test the mounting of the meinspring of the Short & ‘Mason type of altimeter. ‘There are two kinds of test in common use, the general test and the special test for an oxperimental instrument. ‘re fanaa cbt ating appa, ‘Tho general test is tho ono always given if no.other is explicitly requested. It affords data for deducing approximate values of the proper correction under any given condition of use. A modification of the general test is tha so-called short test on service instruments. This is an abridged form of the general test recommended only when time is not available for theALTITUDE INSTRUSIENTS. 519 numerous stages of that test. ‘The operations involved in the general test require a minimum of eight working days; but the short test cen be finished in two days. In most oases the short test will suffice for the rejection of inaccurate instruments, but not always, end it will not provide detailed corrections. Another modification of the general test is further extension of it in the form of supplementary tests on sample instruments. These tests include deter- mination of the effects of vibration, acceleration, or other aeronautic conditions not covered by the general test. Such observations aro important in considering the suitability of new types ‘or makes of instruments, but are not likely to be necessary in testing or inspecting each individual instrument. - ‘Tho special test on experimentel instruments consists of more precise and complete deter- minations of the errors for instruments of high precision to be used in experiments on aircraft. ‘Theso tests may be made by extending the procedure of the general test, or, better, by taking what is kmown as a flight history test. This can only be done after the fight when the flight conditions ere known. By a flight history test is meant one in which the actual variations of pressure and temperature experienced in ihe use of the instrument on a particular occesion are artificially reproduced in the laboratory. GENERAL TEST. Tho general test consists of four parts, A, B, O, and D. ‘TEST A GPRELDMEARY TES). ‘This consists of six parts: The tapping, shift, and inclination tests; the determination of the correction at tho end of the range, end of the aftereffect, by means of a pressure test; and a finel examination for mechanical defects. Of course, any obvious mechanical defects that are noted upon receipt of the instrument may cause its rejection without going through any of the experimental tests. (Q) The tapping tst is made by repeatedly tapping the instrument vigorously enough to just disturb the pointer, and noting the everage deviation of the pointer reading from its mean position. Four or five teps are usually sufficient. @) For tho shift tet tho aneroid is held in a sidewise position and struck edgewise against tho palm of the hand, first with its face to the left, and then to the right. Each time itis after- wards read in a horizontal position. One half the average difference between the left and right readings is recorded, and known as the deviation by shifting, or simply the shift. (3) For the inclination teat the aneroid is held in a horizontal position, read after slightly tapping, and then again read after turning to a vertical position. ‘The difference in readings is recorded and known es the verticel correction or inclination effect. (4) Tho aneroid, having been set to read the altitude as indicated by the pressure of the mercurial standard, is placed in a container and the pressure reduced at once to the lowest point on the scale (the highest altitude). ‘The movement of the pointer is carefully watched during this pressure change to detect jerkiness. After the pressure bas been reduced, the aneroid is allowed to stand a period equal to two-thirds of the time required to ascend to that altitude at a rate of 200 feet per minute. At the end of this rest period, the readings of the standard and of the aneroid are taken end recorded. The pressure in the container is now ‘quickly raised to atmospheric by opening the vacuum system to the air. (6) Five minutes afterwards the aneroid and the standard are again read. ‘The amount by which the pointer fails to come back to the true eltitude is called the aftereffect. This is to be expressed in per cent of tho altitude rangg. (6) The mechanical defects such as jerkiness, parallax, or loose parts which have been. noted during the foregoing test on which may now be found by further inspection are recorded. ‘The design of most aviation aneroid barometers is now such that the tapping, shift, end inclination tests ere unimportant. An error of 60 feet for altimeters and 0.05 inches of mereury for aneroid barometers is excessive. Tf the aneroid passes satisfactorily the six steps of test A, it is put through test B; other- wise it is rejected at this point.520 REPORT NATIONAL ADVISORY CoMMAftER FoR AERONAUTICS. ZEST B (SEALEVEL, TEAPERATORE TEST). This test consists of two steps: (1) After the instrument hes been at the room-temperature of 20° C. for at least threo hours it is tapped and the pointer set to the eltitnds indicated by the mercurial stendard and this reading recorded. It is then placed in a temperature chamber, where the temperature is lowered to —10° ©. and held at this temperature for at least three hours. At tho ond of this period the instrument is again tapped and the position of the index observed, recorded, and compared with the reading of tho stenderd. (2) Tho eneroid is noxt put through the hot test, which is the same as the cold test except that the instrument is hested up to +40° C, and a roading taken after a four-hour interval. From the high temperature tho instru- ment is allowed to cool down to tho room temperature, After three hours it is read again and compared with tho standard. ‘This completes tho temper- ature test at, atmospheric prossure. Nore.—The individual stops of tost “peout®-traonce Eas © ~*~ B ave'pesformed in the order given to +, 22—ulbation carves: 20"C.on inate nade y 8th & tons, 8VOid the condensation of vapor within the instrument whieh would occur if it wore first heated and then cooled, but which is avoided by heating the instrument after the cold test. Care should be taken always to tap the instrument before reading. So 8 crite Feet @ ‘TEST € (CALIBRATION AND DRIFT). (1) After setting the index to the altitude indicated by the standard, the instrument is placed in a container and the pressure decreased at en average rate of 1,000 fect fer each five- minute time interval. Simultaneous readings of the instrument under test and the mercurial standard are taken at intervals of five minutos. Tho differenco between tho two readings is the correstion, which is so given that tho algebraic sum of the correction and the reuding of the instrument under test equals the true altitude. (2) After the pressure in the container has been reduced to correspond to the highest alti- tude indicated on the scale at the above rate, itis held at this value for five hours. At the end of this period the container is tapped and the instrument read, and the correction obtained as before from the mercurial stendard. Ex- treme care must be taken that the temperature of the instrument does not change during the test, and that the pressure in the container be exactly at the same value at the beginning and tho end of the five-hour period, so that the hysteresis effect is eliminated, Much fluctuation in the pressure during the five hours is decidedly to be avoided. ‘The maximum correction of the aneroid barometer reading is noted, and if desired a calibra- tion curve ean be drawa, with corrections plotted against readings, and attached to tho report. Figures 22 and 28 are typical calibration curves, figuro 22 for an altimeter and figuro 23 for an aneroid barometer. ‘The drift is the quotient, in per cent, of the change in the correction found after five hours at the high-eltitude range divided by the true range. ‘The average devia tion is the average of the deviations of the celibration curve from the best representative straight line, FS +10 10 correction = Inches of mercury B29 272525 Reoding» Inches of mercury ‘ro, 28—Cabtion nev fice pete ners.ALTITUDE INSTRUMENTS. 521 {TBST D (ALIITUDE-TEMPEBATURE TEST). ‘Test D consiste of two additional calibration tests: (1) At a temperature of —10°. (2) Ata temperature of +40°. ‘The corrections are found in the same manner as in test C (1). Straight lines are drawn through the calibration curve at each temperature, including the calibration at +20°C. This, will givo three straight lines. ‘The slope of the calibration curve at 20° C. should always Lie between the other two, the slope of calibration curve at ~ 10° C. being greater and at +40° C. being less. This is for curves whose coordinates are in eltitude. ‘The slopes are determined in per cent for the curve, and the difference between the slopes for the lines for +40° C. and—10°C. is noted as the change in scale value. ‘This value must not exceed 5 percent. The intermediate slope at +-20° C. is of value in this test in showing the regularity of the shift with temperature and as a check. For aerobarographs or other instruments reading in inches of pressure, instead of in feet of altitude, the same methods are followed throughout tests.A, B, C, end D, readings being taken at each inch of mercury pressure changes instead of each 1,000 feet. ‘The pressure is changed at the rate of 1 inch of mercury pressure every five minutes instead of 1,000 feet every five minutes. ‘The four tests are purposely given in the order named. Defects due to poor workmanship can usually be discovered before the instrument is calibrated, and if the defect is serious the in- strument is immediately rejected, thus seving time and lebor. The temperature test B is given before the calibration because more instruments fail on this one test then on any other. More- over the instruments are thus given a chance to rest after being strained by undergoung test A. ‘This is an important factor which should not be overlooked. The instruments must be allowed at least 24 hours’ rest after being subjected to a large pressure change, in order to obtain accurate results. Otherwisé, on account of elastic fatigue, the instrument will give a false reeding which may be different from one obtained after « sufficient period of rest. ‘The short test for service instruments differs from the general test given in the following respects: (1) Tests A and D are omitted altogether; (2) test B is made only with cooling instead of with both cooling and heating; (8) the drift observation in test C is omitted and replaced by an observation of the eftereffect. The instrument in this test is held et the low pressure for two hours only instead of for five hours. It is inherently impossible to make the tests A, B, C, and D in less than about seven or eight, days since each instrument must be ellowed a period of time between each stage of the test to recover from the elastic fatigue set up in the metal by virtue of the test itself, and it is preferable that at least two days should elapse between each part of the test. The result of cutting down that time allowance would merely be to produce deceptive figures, which will not reproduce under sviation conditions. In the supplementary test on semple instruments « vibration test lasting for one week is given. The instruments during this test are mounted on the vibration apperatus previously described. During this time observations are occasionally teken to note both the amplitude of oscillations of the pointer and whether any movement of the dial has taken place. Afterwards the instruments are again calibrated. Observations are also made on any other suspected sources of error. In addition to the general test, additional tests are made. on barographs that are to be used for competitive altitude records in order to determine the fitness of the instrament for this special purpose. SPECIAL TESTS ON EXPERIMENTAL INSTRUMENTS. Instruments to be used in experiments on airoraft are carefully readjusted to give the least possible errors, and cslibration curves are then determined. The flight history test referred to above is also given for such instruments when the actual flight conditions are known. Only by such a test can tho offects of elastic fatigue and temperature lag be properly determined.522 REPORT NATIONAL ADVISORY COMMITTEE FOR “AERONAUTICS. RESULTS OF INVESTIGATION. ‘One of the most urgent needs of the aeronautic instrument manufacturer during the recent war was the development of reliable testing apparatus. ‘The first problem was to design a mercurial barometer which could be easily transported and which could be read without tho necessity of making numerous settings and applying corrections such as are roquired with ordinary laboratory barometers. Considerable experimenting was done with different. types of mercurial barometers that were constructed both by the menufacturers and the Bureau of Standards, and their different characteristics were studied in detail. An interesting problem in this connection was the equipment of the fixed cistem types of mercurial barometers with ‘an eltitudo scale, It was attempted to construct a fixed cistern instrument with an equally spaced altitude scale so that it would be possible to use e vernier with this scale. This idea was aKendoned because the investigation showed that the advantages gained by the use of the vernier was offset by the necessity of a very much cramped scale which proved to be a docided disadvantage. - The temperature corrections to fixed cistern barometers were also studied in detail. ‘It was found that the temperature correction in altitude is approximately constant for all altitudes in the ordinary fixed cistern instrument. As a result of the above investigation, a portable standard was designed by the Bureau of Standards, satisfying the conditions that the instrument does not have to be set for the level of the mereury, and for a considerable range of temperature from room temperature it is not necessary to apply @ temperature correction. Another important problem was how and where the routine tests could be modified and still cause the rejection of all the defective instruments. In this connection an investigation, ‘was made to determine how many instruments would pass the “ temperature test at sea level” (test B) and not pass “altitude-temperature test” (test D). ‘Tho resulis showed that {dst D could be omitted for the ordinary service instruments, but this procedure is not recommended for an instrument that is to be used in the performance testing of aircraft, A series of tests was made to determine the shape of the curve, when the slope of the calibration curve was plotted against temperature. ‘The Bureau of Standards carried out an investigation on the thermomotric lag of the various types of instruments for the purpose of determining how long an instrument should be held at a certain temperature during the temperature tests. Tt was also necessary to devise tests to prove the fitness of the instrument to withstand the vibrations experienced under actual conditions of use in aircraft. With this in view, vibration apparatus was designed. .To find out thé effect of vibration the vibration of the pointer was noted; #lso the loosing or movement of any of the parts of the instrument and the effect of the vibration on the calibration eurvo, ete. Defects caused by vibration wore found to be comparatively infrequent. The most important investigation undertaken was a study of the elastic properties of sneroid diaphragm capsules, For convenience these elastic propesties may be divided into three phenomena: —~ Drift, which is the change of displacement, under a constant load, Hysteresis, which is the excess of displacement with load decreasing over the displacement, at the same Ioed with load increasing. Aftereffect, which is the residual displacement at any time after removal of lond. ‘A great number of drift experimenta were made to determine the shape of the drift. ourve, . e., the increase of displacement plotted against time, the variation of the magnitude of the drift with the speed of loading, end the variation of drift. with the load. ‘The improvement of the amount of drift in aneroids submitted to the Bureau of Standards from year to year has been studied, and a curve showing these results is given below (fig. 24). In this curve the average per cent of drift of the eneroids submitted during the year is plotted vertically against the year horizontally. ‘The slight increase in the average drift-during 1019ALTITUDE INSTRUMENTS, 523 and 1920 is due to the fact that several of the instruments submitted were manufactured previous ta the war. ‘A detailed inquiry was made of the hysteresis of various types of instruments and the rela tion of hysteresis to drift was investigated. A curve showing the hysteresis in several instru- ments will be found in Part IT of this report, under precision eltimeter design. ‘While studying the elastic properties of a D-spring type altimeter e detailed investigation of the mechanical errors of this instrument were made. ‘The greatest mechanical error was found to be the method of clamping the minspring to the carriage of this instrament. A very small amount of slipping which occurred here was greatly magnified by the multiplying mechanism, so that the error became large. This error ‘was entirely eliminated by substituting a steel support for the brass one, ordinarily used and preventing the spring from slipping in this steel support by set screws. Slipping of the kmife-edge was found to be : responsible for a similar error. 7 Still another problem in the construction rates. : of the instrument related to the delicate chain ¢ OS ah re in the indicating mechsnism. A device was §¢ constryeted to study the endurance of various types of chain. ‘These tests showed that both § gold and steel chains would far outlast the life =| of eny instrument, but that the gold chain aa ag showed better performance when subject to the a adverse atmosplieric conditions of actual use. 9 L Phosphor bronze strips showed good endurance 88885 88 g and performance when great care was used in ee attaching them in the instrument. Foy 2i-aAvenp dt nod ctl he Bra ef For the performance testing of aircraft itis often very desirable to have an instrument reading in pressure instead of altitude. Most of the aneroids reading in pressure are not suitable for aircraft. ‘To meet this demand the bureau converted several altimeters to pressure instruments, equipping them with scales graduated in inches, millimeters, or millibers. This work led to an interesting study of the multiplying mechanisms in these instruments. Closely connected with the conversion of eltimeters to « pressure basis was the extension of the range of barographs. This wes accomplished by fitting an external spring to the instru- ment so as to increase the stiffness of the system for a given deflection of the box. This creased the pressure range (altitude range) for the same deflection. Special charts were made for these instruments. ; ‘An optical method of testing the regularity of the motion of the drum of a barograph was developed. This consists of putting a sensitized sheet on the drum of a barograph and allowing a very fine line of light to strike the drum parallel to the axis of rotation. If the motion of the drum is uniform the paper upon development should show uniform exposure, if the motion is nonuniform the exposure will show a series of light and dark lines—light when drum is rotating tast and dark when slow.REPORT No. 126. ALTITUDE INSTRUMENTS. PART Il PRECISION ALTIMETER DESIGN. By Jou B, Prrersow ano Jou R. Freeway, Ja, SUMMARY. In this pert the general principles of wltimeter design are discussed and applied to the con- 1 struction of a:large open-scale precision altimoter, tho development of which wes undertaken in accordance with a program epproved by the National Advisory Committeo for Aeronautics. ‘At the beginning of the discussion date aro shown indicating the amount of error due to imperfect elasticity of ordinary eneroids, sowing the great need for improvement in this par- ticular, On the same diegram the hysteresis curve resulting from the finished procision alti- meter is plotted, showing the notable improvement secured. In fact, by tho uso of a stiff steel spring of special quelity coupled to a diaphragm made of ordinary brass but sufficiently thin to contribute practically no elastic reaction, the familiar hysteresis error has actually been made negligible in comparison with the ordinary mechanical errors of the movement. In conclusion, suggestions are made regarding compensation of the design for minor sources of error. Instruments of the precision altimeter class are intended primarily for use in performance testing of aircraft and as working stendards for reference in laboratory tests. INTRODUCTION. ‘The readings of an aneroid barometer are linble to orrors as great as 4 per cent oven after the readings are corrected for the temperature of the atmosphere. In several special cases where accurate determination of pressuro or altitude is desirable if not indispensable, the or- inary commercial altimeter does not have the desired accuracy. It was the purpose of the authors to fuldll this need by first developing « theory of aneroid design and then chocking the results by experiments on an instrument constructed accordingly. Several cases where accurate instruments are essential are: (1) In the performance testing of aircraft; @) In landing at night or in fog; (3) In acrinl mapping; @ In bomb dropping; and (8) As secondary standatd for laboratory or feld use. In performance tests the determination of pressure is more important then that of the altitude. Either an instrument reading in pressure units or an altimeter with a fixed dial should be used. For lending at night, or in fog, the pilot could receive by radio from the landing station the barometer reading: at the station. Knowing this pressure and having e relinblo aneroid, the pilot could descend to this pressure level, i e., ground level, as indicated by his ancroid, without danger of crashing, ‘The necessity for accuracy here is easily seen. Too large an error may be fatal to crow and passengers. 1mALTITUDE INSTRUMENTS. 525 Aerial mapping and bomb dropping are the only uses stated here where the temperature of the air need be eccounted for accurately in the determination of altitudes. This being the case, temperature errors due to deviation from the altitude formula will usually overshadow errors in the pressure determinetion, but ecourate results are desired, and instrumental errors should be climinated when itis practical to do so. An accurate instrument graduated in pressure units could be used to advantage as # seo- ondary standard for the calibration and adjustment of altimeters in the laboratory’ and on the field. For this purpose the instrument should indicate the pressure directly with nearly the accuracy of s mercurial barometer. The ordinary commercial instrument is not satisfactory for this purpose, since its readings, owing to imperfect elasticity of the diaphragms, depend on the rate of change of pressure to which it is subjected. A secondary standard of this type is especially needed for use at aviation fields where mercurial instruments which require - ais careful adjustment and the uso of correction & ,| 2" “beat. Onmerseinstreme tables are very inconvenient. E77 precision altimeter net CI ‘The most difficult error to eliminate or 1 = correct for in an aneroid barometer is the § ,, { i elastic Ing or time effect. Figure 1 shows the oe hysteresis in commercial altimetens. Curve B : is the best instrument of a group of epproxi- LV mately 10 American and foreign makes which $4 were tested. A hysteresis error of 24 per cent 4 15 260 feet oa « 10,000‘ tnstrement, or 500 feet on a 20,000-foot instrument. ‘The 8904s aoe cause of this hysteresis is imperfect elasticity Unite ober Cot ne range of the spring or disphragm, usually due to & mo,1-raamtectecec and hae feet inthe atest stress too high for the material used. Also "eesetyermisute. Tetineatbetop was 9 minute, the mechanical support of the spring and the knife edge which fastens the diaphragm to the spring were found to be sources of error. ESSENTIAL PRINCIPLES OF PRECISION ALTIMETER DESIGN. ‘The design of an altimeter is necessarily such that tho finished instrument is a compromise among the many desirable features. Some of these desirable features are: (1) A large diaphragm, so as to get sufficientwworking foree to operate the mecha ‘nism properly. (2) A spring with low maximum stress, stiff relative to the diaphragm. (3) A large deffection combined with a small multiplying ratio. £ After the outside dismeter of the instrument has been decided upon, the problem is to put the best possible mechanism in the space available. DIAPHRAGM DESIGN. [EFPDOTIVE AREA. ‘Tho effective area of a diaphragm is defined as the ratio of the distending force applied at the center to the pressure required to produce this force. ‘To calculate the effective ares of a diaphragm, let us assumo that a section of the diaphragm from the center to the outside rim, ABCD (fig. 2), acts as a flexible cable with a uniform stress throughout « givan section. Awan angular increment including the area ABCD. ‘P= the atmospheric pressure in pounds per square inch. sngle of departure of the diaphragm from face of center plate.526 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. Using the notation of figure 2, the aren of the section ABCD is } (R*4+2Rr)4u. The load on this areais 4 (2?-+22r)du P, supported partly at the ceniter and partly at the outside rim. The moment of theload on the area ABCD about the tangent to the circle at BC is (R’rdu+}240)P. Divide this moment by R and wo get the part af the load supported by tho center, approximately Gr+4i)40P. Now the total pressure falling on the rigid conter of which AOD is a sector is transmitted to the spring and has there- fore to be added. The load at the center due to the pressure on the triangle BOO is @#"+3Rr+4r)eP. The total load on the conter is DY GE SB td aeP = GRE Rett eP. The effective ares then, is, by definition, Cee) GR Retr) ¥.2tonponen daotcers. ‘The ratio of the effective ares to the actual ares is p+ Rete 7 Repke which may be expressed in the form This ratio varies from one-third when r is zero to unity when R is zero. ‘The foregoing anelysis is an approximation satisfactory for relatively small values of the inner radius r. Ping STR. ‘The maximum fiber stress in the diaphragm occurs where it is fastened to the rigid center. Considering the area ABCD again (fig. 2), the vertical load on the center due to this area was found to be (}Rr+4F*)AwP. Now we denote the angle of departure from the center (fig. 2) and consider the radial tension in the diaphragm material to be the resultant of the vertical and a horizontal pull. The total radial tension in the section AD is— ‘Tho aren of the section in tension at the canter is ru, t being the thicknees of the materiel ‘The fiber stress is then— po GFet4P) Ae ride sin & Since the area supporting this load at any other placo in the diephragm varies directly as the distance from the center, the stress at any point is Gt ie where r; is the distance from the center to the point where the stress is to be calculated. This is for a smooth disphregm.ALZITUDE INSTRUMENTS. 527 In a corrugated diaphragm the maximum stress ocours at the bottom of the corrugation, ‘and, using the sysabols of figure 3, is . Pas (1+) ~n Ete) (+P) rare For the proportions shown in figure 3, 1+57=16. In order that the maximum fiber stress shall not exceed a certain limit, diaphragm heving corrugations as in figure 3 should theo- rotically be 16 times as thick as a smooth diaphragm of tho same radius and angle of departure from the ae go In tho design of precision altimeters the dis- a phragm is the element which limits the general size a of the mechanism. In the Bureau of Standards pre- Sane, . cision altimeter No. 1, which is described in deteil later in the paper, the diaphragm diameter is about 8 inches. Tt is belioved, however, that diametar of 5 inches will give the proper relation of diaphragm area to spring stress and deflection. The total area of the diaphragm is then 19.6 square inches. Substituting in the above equations, the effective ares is found to be 8.6 square inches, or 44 por cont of the total area. This area is assumed to be constant for all positions of the diaphragm. ‘At 147 pounds par square inch atmospheric pressure the load which the diaphragm puts on the spring is 8.6x14.7=126 pounds. At 30,000 fet the pressure is about one-third of thet at sea level. ‘The load on the spring is 8.6x4.9=42 pounds. It is, therefore, required that the spring be designed to carry maximum load of 126 pounds, with a working range of 42 to 18 pounds. Teis a matter of preference whether the diaphragin be corrugated or smooth. As developed above, the metal for a corrugated diaphragm should be several times the thickness of metal for a smooth diaphragm of the same size. Phosphor bronze and nickel silver are good materials for diaphragms. ‘The stress should not exceed 12,000 pounds per square inch for either of these metals. RIPE COMPENSATING DIAPHRAGM. If the altimeter diaphragm hes a variable aret, as shown diagrammatically in figure 4, it has two advantages over the constant ares diaphragm. In the first place the area increases ‘as the altitude increases. This tends to make a straight deflection-altitude curve, or, with a con- stant multiplying ratio, a uniform altitude seale. ‘The other advantage is the drift compensating feature. If the pressure on the instrument is re- duced until the pressure corresponds to & chosen altitude, and is held constant at this value, the reeding then increases gradually, owing to drift in the spring and diaphragm. With the variable ares diaphragm, this drift causes an increase in the effective diaphragm ares, thereby augmenting the load on the spring which, in turn, tends to diminish the drift. ‘This has been verified experimentally by the fact thet the spring used in the precision altimeter, described below, shows less hysteresis when coupled with the diaphragm as used in the instrument than does the spring alone when tested by defleoting it with weights. ‘A possible disadvantage of the variable ares diaphragm is the lisbility to a change in the calibration caused by stretching of the diaphragm. Further experience is necessary before the seriousness of this possibility oan be determined. ‘ro. 4 Dat competing aphrgm.528 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS, ‘SPRING DESIGN. Tt can be proved that the hysteresis in an aneroid barometer may be reduced to a very low value if a spring with good elastic properties is used in conjunetion with a relatively very floxible diaphragm, although the diaphragm itself may havo poor elestic properties. ‘This principle has been applied in the construction of the altimeter described in this paper and may be demonstrated as follows:* If two springs with stiffness 8, and S,, respectively, are connected together es ‘Sy "in figure 5, the stiffness of the system to 2 foree applied at Ais S=S,+5, (Stiff ‘ness is defined es the amount of force epplied per unit of defleation.) Now, if h, and fh, axe tho hysteresis values of the two springs, respectively, for a certain range of deflection, the hysteresis of the point A for the system is— 78 SiS ‘ream heh, Bt, Y . If; is the stiffness of the epring and 5, is the stifiness of the diaphragm; e.g. mated that in the Bureau of Stendards precision altimeter No. 1 $=0.99 and S'=0.01. If the hysteresis in the spring for the maximum range is 0.02 per cent and the hysteresis in the diaphragm for the corresponding range is 2 per cent, the hysteresis of the combination is is eati- 99 951 10.02 95 +2-0; 49 =0.04 per cent. COMPOSITION OF THE STEEL AND PEHDMSSIBLE STRESS. Soveral investigators, in particular Bairstow,’ and Smith and Wedgewood * have studied the possible relations existing between hysteresis in stecl and the fatigue strength. ‘They have shown thet the limit of proportionality as ordinarily determined can not be taken as a ori- terion of the limiting stress bolow which there is negligible or zero hysteresis, but that the fatigue limit as determined by repeated stress is the limit below whickt there is no measurable hysteresis. Further, the width of the hysteresis loop increases with increasing stresses beyond the fatigue limit and for a stress range greater than the fatigue range the width of the hysteresis loop increases with increasing number of repetitions of stress. ‘Using these facts as a basis, it is readily seen that the permissible stress in an aneroid spring, or in any steel spring where true elastic reaction is necessary, is the fatigue limit of the steel used. This futigue limit for good grades of commercial alloy spring stecls is approxi- mately 25,000 to 30,000’ pounds per square inch fber stress. However, since the width of the hhysteresis loop is very small for stresses somewhat above the fatigue limit, and because of the relatively very low number of stress alterations an aneroid spring would undergo during its e, compared with the number required to cause failure from fatigue, a maximum fiber stress of 50,000 pounds per square inch is probably not excessive for heat-treated alloy steels. ‘This is especially true in an aneroid spring where te stress is never reversed (tension to com- pression) and there is a consequent tendency for the spring to adjust itself to the range over which it operates. - ‘Extensive experimenting was done in the attempt to make a suitable spring for the pre- cision altimeter No.1. * ‘The best spring obtained was of a special nickel-silicon steel having a yield point of 276,500 pounds per square inch. Tnereasing the maximum fiber stress of this spring from zero to 100,000 pounds per square inch and back to zero in 20 minutes, the maximum width of the hysteresis loop was 0.3 per cant of the total deflection. To ba on the safe side and to permit the use of ACD, Hear. Theory of6itaraof hale yatent. oamal afin Waingon Amdeny often, VoL V3,p.06, 1h The prin cz pn il epi sop oe Cpe pop Me Ses ta of pt wc a Dalr, Lewanls Press. 1, 20,9. 38 althnd Wedgoncod. foun. tron iol nt... 28,1985ALTITUDE INSTRUMENTS. 529 commercial heat-treated elloy steels, 50,000 pounds per square inch is chosen as an absolute maximum permissible fiber stress. ‘The nickel-silicon steel on range of 60,000 pounds per square inch should give a hysteresis of only one-tenth of 1 per cent. If the range of the instru- ment is 20 inches of mercury, the maximum width of the hysteresis loop is 0.02 inch, or 30 fect if the instrument is an altimeter. SSUAPE, STIFFNESS, AND DEFLECTION. ‘The spring should be shaped so that as near as practicable there is « uniform stress over its entire length. This method decreases tho stiffness but increases the deflection without increasing the maximum fiber stress. Tt is desirable to have the spring straight at either tho middlo of its working range or at sea-level pressure. To accomplish this, the spring is made so that its free position is curved. If it is to be straight at the load which is the middle of the working range, its free shape should be the reverse of the curve teken by a similar straight spring when its total load is applied. ‘The following method may be used to calculate the deflection of a leaf spring the width and thickness of which varies with the position along the length of the spring. ‘Figure 6 represents a cantilever spring which is the same as one-half of the altimeter spring. The cantilever is fastened at A, and this corresponds to the middle of the altimeter spring. ‘The notation used is: : Readius of ourvature of the spring. E=modulus of elasticity of the steel. Inthe moment of inertia of the section. M= the moment caused by the load. P=the load, concentrated et the end. b= the width of the section. t=the thickness of the section. z=the distence from the end. Dathe deflection at the end. forthe fiber stress. ‘a= the length of an increment for graphical computation. ‘The radius of curvaturo at any point elong the length is _BI_ Eve En Pe B. ‘The angle « is then a/R, ete. ‘The deflection, 2,= 255% ne (sina 482 G49) ‘The end deflection is D=a2[(sin a,) +sin (a, +a,) +--+ -- dsin(qyt--- a] ‘This formula is used for coloulating the shape of the free spring so that it will be straight ‘at any required load, and elso for calculating the stiffness and deflection. It has been found by experiment that tho stiffness for such deflections as here experienced is practically constant. Complete calculations for the spring to be used with the 5-inch diaphragm previously dis- cussed will be given. The modulus of elasticity of the steel is taken as 29%10* pounds per squere inch. The correct thickness for 50,000 pounds per square inch maximum fiber stress is found to be about 0.095 inch. For the shape calculation the load et the end is 42 pounds, or one-half of the diaphragm pull at the middle of the working range. The length, 33 inches, i divided into six parts, each 0.54 inches long. sm167—23—84580 REPORT NATIONAL ADVISORY COMMITTEE FOR AFRONAUTICS. EI_Ee _}_20x10°x,000857 _ 49 4b “Te fop 27 iaxds 49-8; o.o1s | O0n5.} 0.0115 oo | cu {cone | “eas ca oo |** .osi8. 0318 087 | He | ease ‘| fala coe | ‘Tho total defection is 0.095 inch. To get the shape dimensions given in figure 6 each de- flection is subtracted from the total deflection. Po. 6g aeceton, 2x42 The stiffness, += 226%.-890 pounds per inch. ‘The deflection for the working range is p7-095 inch. ‘Tho fiber stress— ss AE 6X 3.25 x 63 : SM. OX225%8 _ 49,500 pounds per square ich. One-half of the maximum load of 126 pounds is used in calculating the fiber stress. MULTIPLYING MECHANISM. Since the area of the diaphragm and the stiffness of the spring remain constant, the dia- phragm deflection will bear linear relation to the pressure. The problem is to design multiplying mechanism such that the movement of the end of the pointer bears a direct linear relation to the movement of the diaphragm. Fro. —Multpying mecha, Figure 7 is'a representation of # lover mechanism. The sector shaft and the main shaft are perpendicular end in the same plane. ; Let a, b, and ¢ be tho lengths of the first, second, and third lever arms, respectively. Let A be tho angle between a and the horizonital. Let B be the angle between b and the vertical. Let be the angle between ¢ and the plane drawn through the axes of both the main shaft and the sector shaft,ALTITODE INSTRUMENTS. 581 Let D be the angular defleetion of the pointer. Let z be the vertical movement of the diaphragms (r= 0 when the arm a is hori- zontal). . Let 2 be the multiplying ratio of the sector to the pinion. Assume that when the lever a is horizontal, b is vertical, and cis in the same plane as the sector shaft and the main shaft. ‘Thet is, when z=0, 4=0, B=0, = 0. Now the locus of the point of contact between the lever arms b and c is a straight line, perpendicular to the plane in which the main shaft and the sector shaft were placed. We will use m es a symbol to denote the distance of the point of contact from this plane. m= tan B and m=e ten C ve ten O=b ten B dut Basin-*Z F int see tan Omd tan sintZ om tan (2 tan sin+2) Do R tant (2 tan sir2) If we express this equation as a series it appears that D comes very close to being exactly pro- portional to z when ® is sbout 1.25. As developed, the multiplying ratio is approximately constant for all positions of the diaphragm. It follows therefore that the pressure scale will be uniform. The shape of the calibration curve may be adjusted by changing the ratio of the crossed. lever arms, and the seale value, or the general direction of the calibration curve, by adjusting the length of the arm making contect with the top of the diaphragm. ‘These adjustments should be made in the order mentioned, for the first changes the scale value but the lest does not change the shape of the calibration curve. ‘TEMPERATURE COMPENSATION. It is not practicable to formulate exactly a method of temperature compensation until the nature and magnitude of the temperature errors are determined experimentally. In generel, the introduction of one or more bimetallic bars somewhere in the lever mechanism will com- pensate the instrament. In a bimetallic bar composed of two metals of different temperature coefficients of expan- sion, there is usually within eech metal neutral surface which receives no stress, the metal on one side of this surface being in tension and the metal on the other side being in compression. Tt ean be proved thet through a lenge range of temperatures these neutral surfaces do not change their position in the bar and, therefore, the distance between them remains constent. When o ber, straight at one temperature, tales a curved shape at another temperature, this curve is the arc of « circle, and the radius of curvature is the same for any position along the length of the bar. Starting with these assumptions, formule for calculating the deflection of any bimetallic bar have been developed, leading to the conclusion that such a bar will give maximum deflection when the steel strip is approximately two-thirds os thick as the brass strip in the usual ease of the brass and steel combination.532 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. [BEPECT OP AIR IN VACUUM BOX. Another possibility of temperature compensation is the adjustment of the amount of airleft in the vacuum box. Less than one-half millimeter of air in thebox has noeffect on theoperation of thb instrument, and it is therefore never necessary to exhaust to e high vacuum. For temperature compensation 5 millimeters is the least emount which will have an appreciable effect. ‘The box of the precision altimeter is exhausted to 0.2 millimeter. POSITION ERROR AND BALANCING FOR VIBRATION. ‘Tho most serious.vibration effect is caused by « vibration etrright anglea to the plane of the diaphragm. This is because it is not practical to balance the spring and diaphragm against, other parta, ‘The position error is a maximum ‘when the face is down. ‘To minimize position errors, the instrument should be calibrated and mounted in the airplane with the dial in the same relative posi Both sector and pinion should bo balanced about their axes. By referring to figure 7 it, can be seen that an engular acceleration of the instrument in a clockwise direction tends to cause the pointer and the sector to move sn e counterclockwise direction relative to the rest of the instrument. Now, they both ean not. move counterclockwise because they are geared LUGE Tunene, together. If the moment of inertin of sector Borat Sanger rc ears the same ratio to that of the pointer and pinion assembly as the ratio of the number of teath on the sector to that on the pinion they will be balanced and the pointer will keep its position on the correct graduation. ‘To get this ratio of the moments of inertia the sector should be made heavy and the pointer light, THE BUREAU OF STANDARDS PRECISION ALTIMETER NO. 1. ‘This instrument was developed by the Bureau of Standards for the Army and the National Advisory Committee for Aeronsutios, based on ‘the theoretical considerations given above. In Cy it are eliminated to a largo oxtent the errors Rak wl & commonly found in aneroid barometers, which have also been discussed in detail in this paper. Referring to figure 8, a Aexible diaphragm. (2) is coupled to a stiff mainspring (2) by means of yoke (5), the two bolts of which screw into the center plate of the diaphragm. ‘The main- spring is swung on two thin flexible springs (8) from spring supports (4). This method of support permits the free leteral movement °—' "yiisaeis mousonts offec? of the end of the spring when deflected. A. short pin mounted at the end of a short arm. from the spindle (6) fits into a conical bearing soldered in the center of the diaphragm plate. Deflections of the diaphragm cause (6) to rotate, This motion is multiplied by upright (7), which is connected by rod (8) to tail of cam (9). A hairspring mounted on the same shaft es cam (9) maintains a slight pressure on the conical bearing. Cam (9) is made with a ‘varying radius so as to facilitate the adjusting of the lever system to give an equally spaced altitude scale. ‘The movement of this com is communicated to a sector (10) by means of ‘small roller bearing on the end of arm from the sector shaft. This sector meshes with pinion (11) and rotates the hand (12). The roller bearing is held on the cam (9) by the tension of the hairspring (13). & ¢ Correction in feet ‘a 0-Prslon tae clan re,ALTITUDE INSTRUMENTS, 588. Figure 9 is a calibration curve of this instrument. It is important to notice thet the differ- ence between the up and down readings is only 15 feet and is very small compared with that of an ordinary service instrument. ‘The instrument émbodies the drift compensating feature. As previously stated, thero is some doubt es to whother an instrument, with this drift compen~ sating feature will hold its celibration, but this particular instrument has shown no change. ‘This point is under further investigation, however. Preliminary tests indicate the temperature error of this instrument without bimetallic bars to be smell. ‘The suecessful development of a precision instrument of this type is dependent on very care- ful workmenship. The authors were fortunate in having the assistence of Mfr. F. Cordero, of the Bureau of Standards, in the construction of this instrument.REPORT No. 126. ALTITUDE INSTRUMENTS. PART U1. STATOSCOPES AND RATE-OF-CLIMB INDICATORS. By Avmzmon H, Means, INTRODUCTION. Statoscopes:are used for indicating when aircraft are maintaining « constant altitude, or for quickly determining small changes of altitude. ‘The ordinary altitude instruments, the alti meter and the barograph, aro of little use for this purpose. ‘They are too slow in action and not sufficiently sensitive. Few altimeters or berographe will indicate with reliability a change of altitude of 20 feet. The statoscope is especially useful in the navigation of balloons and dirigi- bles, since it shows immediately when the aireraft is ascending or descending. This gives the navigator warning before the airship has hed time to gain an appreciable verticel velocity; and thus avoids the necessity of making wasteful adjustments of either the gas or ballest. Stato- scopes, of the bubble type, can also be used to indicate approximately the rate of ascent SS or descent of an airplane by determining th rate at which the bubble breaks. ‘Rate-of-climb indicators have the advan- tage over statoscopes of giving at once the ep- proximate rate of climb, without time obser- vations and in most cases without subsequent calculations. ‘This is advantageous in helping the pilot to attain his maximum climbing Sera speed, for example in aircraft performance tests, since he has only to observe the instru- ment and so manipulate the controls as to get'the meximum indications. In the landing of be loons and dirigibles it is also very important to know the rate of dascent, ‘The Bureau of Standards has recently developed a mechanical type of rato-of-ctim cator of such sensitiveness and accuracy as to fulfill the requirements of both statoscopes ant rate-of-climb indicators on nearly ell types of aircraft. DESCRIPTION OF INSTRUMENTS. ‘STaTOSCOPES. Bubble statoscopes—American.—The indicating bubble type, as shown in the following photo- graph (fg. 1) and diagram (Gg. 2), consists of o thermally insulated air chamber with an outlet to the external air. To this is attached a small special radium-illuminated curved glass gauge or manometer. ‘The curvature of this gauge is varied according to the desired sensitivity to change in altitude. In this gauge is placed a very small amount of liquid, whieh should have low density and low vapor pressure, Its viscosity should not be excessive at a low temperature. At each end of the gauge is blown e glass trap which prevents the liquid from escaping either into the air chamber or to the outside air. When the pressure is changed in any manner the 53ALTITUDE INSTRUMENTS. 535 liquid moves toward thet end of the gauge which is at the lower pressure. The enlarged end of the gauge causes the liquid to form a bubble (hence the name bubble statoscope). ‘This bubble is now pushed still farther into the enlarged section of the gauge and breaks, allowing the air to flow past, thus equelizing the pressure between the inside and outside air. The liquid now flows back into the center of the gauge and forms an indicating medium again. ‘The above eycle of opera- tion continies as long as there is a change in pressure. 1s 2 Diagam of Custer buble staocope Fie 4Brtleh Wg bebe sttonere ‘The sensitivity is such as to produce a movement of the bubble equal to one of the arbitrary scale divisions for a pressure change of 0.02 of an inch of mercury. ‘This corresponds to “Bn Sicha statcope aphrge,maltiplyng mechanism ad change in altitude of 20 feet at sea level or 26 feet at an altitude of 10,000 feet. This type of instrument hus been the most extensively used of the various types of statoscopes. Bubble statoscopes—British —The essential details and operations of this statoscope (fg. 3) are the same as in the American instrument. Considerable care is taken in the thermal insu- lation of the air chamber, which is a Dewar flask jncased in'wool felt. The sensitivity of the instrument is such that, e change in pressure corresponding to 2 or 3 feet of altitude at sea level is indicated by this instrument.586 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. Mechanical statoscopes—American:—The indicating pointer type shown by photographs (Gigs. 4 and 5) and diagrams (Sg. 7), consists of ‘a cylindrical metallic sir chamber (1), one end of which is a very thin, flexible, corrugated, metallic diephragm (2). Deflections of this diephragm are. indicated on the dial of the instrument by means of pointer (8), which is actuated by the multiplying mechanism in the following manner: The motion of the dia- phragm is transmitted to the multiplying mechanism by means of the upright soldered to the center of the disphragm. This is in contact with one arm of the bell-crank shaped lever (3) to the other arm of which is attached a connecting link (4). ‘The other end of this connecting link is attached to a small cam (5), on the shaft of which is mounted the pointer (6). ‘There is an outlet (8) to the air chamber to which is attnched smell rubber tube. The knurled-headed screw (9) is used to adjust the zero setting of the pointer. To operate the instru- ment the opening in the air chamber is closed by pinching the rubber tubing. If now the external air pressure on the instru- ment is changed, the diaphragm deflects, duo to the difference in pressure; outward if the pressure is reduced, indicating an ascent; and inward if ineressed, showing a descent. ‘The dial of the instrument has a luminous arbitrarily divided scale. ‘The value of the divisions is determined by experiment. The instrument shown in the figures ¢ end § was not thermally in- sulated, hence slight variations in temperature cause it to deflect when the outlet is closed, thus making it unreliable. RATE-OF-CLIMB INDICATORS, The katanoscope, chronometric type—This instrument is « modilication of the mechanical statoscope described above. The outlet to the external air in the instrument is automatically opened and closed at regular time intervals (every 20 seconds) by means of clockwork. It has 2 thermally insulated air chamber the diaphragm of which, is made of rubber, oiled silk, or similar materials. Deflections of this diephragm, as in the case of the mechanical statoscope, are indicated on the dial of the instrument through the intermediary action of» multi- plying mechanism, The operation of tho instrument is as follows: Suppose the pressure of the external air to bo changing es in the flight of an air- eraft, The clockwork controlling the suto- matic valve is stated. This closes the auto- matic valve, which traps the air in the chamber. The diaphragm deftects, due to the change of external air pressure, this mo- tion of the diaphragm being indicated by the movement of the pointer over the dial of the instrument. This deflection continues until the valve is opened by the clockwork, equalizing the internal and. external pres- sure. The pointer now goes back to zero, indicating that the diaphragm is back to its initial position. ‘The amplitude of this periodic movement of the pointer is a measure of the rato of change of pressure, from which the rate of climb can be computed. ‘These instruments are fumished with an arbitrary scale, the values of its division in rate of change of pressure (rate of climb) being determined by experiment," From data obtained in the laboratory, a Fa tdasmsiy dour acbastel slate.ALTITUDE INSTRUMENTS. 587 table or chart may be constructed of the indications of the instrument at various rates of climb and altitudes. From this the resdings of the instrument can be converted into rates of climb at all altitudes. Since the instrument indicates only pressure differences, the above table is necessary. A little consideration of the pressure altitude relation shows this instrament will give widely different readings for the seme rate of climb at different altitudes. Leak type of rate-of-climb indicators. —The inherent, disadvantages of the katanoscope have lead to the development of rate-of-climb indicators of the capillary-leak type. These instru- ments all operate on the seme general principle. A thermelly insulated air chamber is closed to the external air, except for a small opening which may consist of either a needle valve or capillary tube. If the pressure of the external air is varying in eny manner, the pressure inside the air chamber will lag behind that of the external air, due to the resistance offered to the equalization of pressure by the leak opening. This causes a pressure difference which is measured ‘io. Rate aeotclin indian and Baran of ‘ia. 8. —Ligua type atelier. ‘Standards rateatinb neater. by some type of indicating mechsnism, either a liquid manometer or the deflection of a sensitive disphragm. The nonturbulent flow through capillary tubes depends.on the viscosity of the air; and since the viscosity is independent of the sir density, it can be easily proved that en instrument constructed on this principle and graduated to indicate rate of climb will have a scale value which will be epproximately independent of the altitude as determined by pressure. ‘The scale value of « rate-of-climb indicator is here defined as the ratio of the true rate of climb to that indicated by the instrament. ‘Models using a liquid manometer—British R. A. E. model —This instrument is typical of all liquid types of rate-of-climb indicators, which consist of on air chamber closed to the air except for aleak or vent. The pressure difference between the internal and surrotinding air is measured by aliquid manometer, in most cases en ordinary U tube. This instrument, illustrated on the following pages (figs. 8, 9, end 10), consists of thermhally insulated chamber (Dewar bulb) (6) closed to the sir except for capillary leak tube (5). Any pressure difference between the tivo ends of the capillary tube due to variation of the external air pressure is indicated by the liquid (7) rising or falling in the specielly constructed manometer tube (4). Both ends of this manometer are blown in such a manner as to prevent the liquid from spilling out no matter588 REPORT NATIONAL ADVISORY COMSUTTEE FOR AERONAUTICS. how the instrument is orientated or to what pressure change itis subjected, a decided advantage over some other types of instrument ‘Thore is attached to the trap (3), by means of a rubber tube (2), a device for flushing the manometer (1). This is operated by pressing a small knob inward, then closing the hole through Pia. 2008. A.B, (Blan) atest indleatee, = lo. t2—Huoand al netr andre are this knob by the finger and allowing the knob to be pressed outward by the spring tension. Tho suction produced. causes the liquid to riso in the manometer tube, thts Hushing it This instrument-hes a rango of rate of ascont up to 1,100 feet per minute and a rate uf descent of 200 feet per minute. The time lag is rather large in this instrument, about 30 seconds. That is the time interval required before the instrument indicates its.true rate of climb when properly calibrated,aurirupe ixsrruatewts. 539 The British Wright Co. rise and fall indicator—The main details of this rate-of-climb indicator, as may be seen by referring to the following photographs (figs. 11 and 12), ere the same es in the previously described instrament. The glass capillary is replaced by one of platinum. ‘The specially designed manometer is replaced by a. U tube. ‘There is a cock attached to the manometer which when closed. prevents the instrument from indicating. ‘The range is from 0 to 2,000 feet. per minute, both for ascent and descent. German balloon variometer (Big. 13) —This 7 instrument is similar to the British’ R. A. E. described above. ‘The main difference is the addition of a filter and dryer for the air before it passes through the capillary leak tube. The manometer is inclined and henee requires only ‘a small pressure difference for a given reading, thereby cutting down the time Ing but at the same time making it necessary to exercise great care in maintaining the instrament level. ‘The air chamber is a Dewar bulb surrounded with cork, which provides exceptionally good thermal insulation. ‘This instrument is made in two ranges from 0 to 600, or 0 to 1,200 feet per min- ute. The time lag is small, about 5 seconds. Sometimes instruments of both renges ere ‘mounted together, the combination being known as a double variometer. Vertimeter (Gg. 14)—This. instrument, of American manufacture, presents a unique de- parture from the general construction of rate-of-climb indicators. The air chamber consists of a large uninsulated streamlined receptacle, which is mounted on the plane eway from the cockpit, where it will maintain the temperature of the free air. The indicating mechanism of this instrument is shown in the following diagram (Bg. 15). ‘The air chamber is attached to the indiceting mechanism by « long rubber tubing, the length and bore of which must be of the proper dimensions or the calibration of the instrument will be affected. The outlet (6) of the indicating mechanism is directly connected to the manometer (1) and a needle valve (), the needle valve taking the place of the capillary leek tubes in the other instruments. ‘This valve is in communication with the surrounding air through the outlet (6). When the needle valve is adjusted so that the manometer gives the proper indications of rate of climb, the valve lock nut (7) of the valve is soldered in position. ‘The trap in this instrument does not prevent the escape of liquid, either into the external air if the rate of ascent is excessive or into the needle valve and the air chamber in the case of too rapid descent. ‘The design of the trap is such that a loss of liquid oceurs whenever the velocity of the latter is great. In the former cese the instrument must be refilled, and in the latter the needle valve must be takesi out and cleaned and the instrument recalibrated. Either contingency is likely to happen in flight, especially if the instrument is tilted during climb or descent. The range of the instrument is from 0 to 2,000 feet per minute for ascent and 0 to 3,000 feet per minute for descent. The time lag is stall,.5 seconds, but the scale is not so open es in other instruments. TT veiinetr an Ameria rated laden540 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. Models using a diaphragm manoineter—Mechariical rate-of-climb indicators differ radically from the liquid type in the substitution of a sensitive diaphragm in the pressure-measuring cle- ment in place of the liquid manometer. Several models of this type have been designed and constructed at the Bureau of Standards. Two designs, known as models No. 2 and No. 3, are in practical uso and aro described below in detail Bureau of Standards, model No. 2: The operation of this instrument may be studied by referring to the following diagram and Meese iche stem EB Eierderacmer’ photographs (figs. 16 and 17). A bank of SHESEE De PEST cen metalic dphragne (0) conte together ‘neicoting turd at the center by metallic rings forms the chamber of this instrument, ‘To this chamber is attached a capillary leak tube (2.) ‘The deflection of these diaphragms is com- municated to the indicating hand attached at (8) through. the multiplying lover (3), by the phosphor-bronze strip (4). This motion is communicated to the arbor (7) by means of another phosphior-bronze strip (4). A slight tension is kept on this system by the hairspring (6). ‘Thermal insulation is provided by an tir jacket around the mechanism. ‘The range of this instrument is from 0 to 2,000 feet. per minute both for ascent and descent, but it can be subjected to all rates that would be experi- ‘enced in the flight of aireraft without damage. The time lag of the instrument is 30 seconds. Burea of Standards, model No. 8: Certain modifications of the above-described model were found desirable in order to reduce the time lag, and to diminish other sourees of error.ALTITUDE INSTRUMENTS. 541 ‘These instruments were incorporated in the design of # new instrument for the Balloon and Airship Division, Army Air Service. Referring to diagrams and illustrations of model No. 3 (figs. 18, 19, and 20), the following are the more important changes: First. The substitution of a large-diameter rigid type of air chamber (1) with thin die~ phragm (8) in placo of a bank of metallic diaphregms connected together at the center by metallic rings, which formed the air chamber in the older model. This modification makes it possible to get a sufficient deflection of the diaphragm with a smaller pressure difference, thus Pig 18 at olin theatre NO cutting down the time lag to a fraction of its former value, and at the same time eliminating the largo inclination error in the previous instrument, due to the mounting and mass of the dis phragms. Second. The use of a glass capillary tube (6) instead of éno of copper. ‘This modification suggested itself when the calibration of model No. 2 was observed to change slightly, which was thought to be due to corrosion of copper tube. ‘Third. The elimination of all levers, facilitating the counterbelancing of the working parts of the instrument. "Fie 19 Rated nde, oa Noa. means of adjusting the calibration of the instrument; and it provides & means of compensating the instrument for temperature. ‘The operation of the mechenism is as follows: Deflections in the disphragm (3) are malti- plied and communicated to the indicating hend (28) through the phosphor-bronze connecting strip (20), which actuates the multiplying pulley (18), which in turn transmits the motion to the pulley on the indicating erbor (26) by means of the phosphor-bronzo connecting strip (21).542 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS, This motion is read on the dial by means of the indicating hand (28). ‘The helical spring (13) serves the purpose of taking up all backlash in the instrument, and also to adjust the zero setting of the indicating hand by varying the tension on the diaphragm (3) by means of the slide (15) and the adjusting screw (16). Tho rangé of the instrument is from 0 to 3,000 feet, per minute ascent to 0 to 3,000 feet per minute descent, but, like the previous model, ean be sub- jected to all rateg up to about 9,000 feet per minute ascont end to 30,000 feet por minute descent ‘without injury, sb that itis almost impossible to damage the instrument by subjecting it to a ‘too great change in pressure, ‘The time lag is about 5 seconds. German mechanical rate-of-climd indicator —This instrument (Bigs. 21 to 24) is similar in many respects to the Bureau of Standards model No. 3, although the two instruments were developed independently. The operation of the instrument is as follows: ‘The movement of the diaphragm is communicated to the pointer by means of a connecting thread (3), which actuates the multiplying lover (6). ‘This lever is counterbalanced by a’counterweight. ‘The Duras of Sod ain nd, ata He, 3-—darman neti lr hae Ne, multiplying arm of this lever communicates its motion to the indicating pulley (7) to which is atteched the pointer (8) and is rotated by means of the winding and unwinding of the thread on the pulley against the tension of the hairspring (9) which is attached to the same shaft. ‘This shaft is also counterbalanced. . The motion of the above mechanism is limited by two stops inserted into the multiplying lever and bent so as to allow for proper movement of the pointer, ‘but preventing the mechanism from being damaged by excessive pressure differences. ‘There is a zero regulator attached. Tt operates by putting tension on a light helical spring (18) which is attached to the center of the diaphragm. ‘This is accomplished by turning the knurled screw (12) placed on the outside of the instrument case, which winds or unwinds thread (14), the other end of which passes over guide pulley (15), thus adjusting the tension on the spring (18). ‘The leak deviea of this instrument is unusuel,. It consists of e tube (1), into which is in- serted a tapering pin. The instrument is calibrated by adjusting the position of this pin inALTITUDE DSTRUMENTS. 548 the tube. Thermal insulation is effected by an air jacket between the air chamber and the outer case. ‘The range of the instrument is from 0 to 1,000 feet per minute, both for descent and ascent. ‘The time lag is about the same as in Bureau of Standards model No. 3. e Magnetic type —While instruments of the leak type and automatic statoscopes represented by the Kkatanoscope constitute the only types of rate-of- climb indicators which have found practical use, various other types of rate-of-climb indicators have been suggested. A magnetic instrument of Dutch manufacture uses & propellerdriven armature re- volving in the field of a permanent magnet. ‘This magnet swings like a pendulum and is always ina verticel position. ‘The rate of climb is determined by the speed of rotation and orientation of the arma- ture in the magnetic field, since the rate of climb is a fanetion of the air speed and the position of the armature in the magnetio field. (COMBINATION STATOSCOPE AND RATE-OF-CLDMB INDICATOR. Tt hes been suggested thet en instrument be designed to function both as a statoscope and a rate-of-climb indicator. One way of doing this would be to provide a valve to close the leale orifice, in which case the instrument can be used as a statoscope. With instruments as sensi- Fis Garman mivotelinb iar. - Fon Geman calectimb ndete, tive as the mechanical rate-of-climb indicator this will not be necessary for héuvier-than-air craft, since these instruments indicate a rate of climb as small es 20 feet per minute; however, we # i 44-—German nieofelia ndae, ase in the operation of lighter-than-air craft an instrument giving much more sensitive indications than is possible with a rate-of-climb indicator is required, and for this purpose a combined instru- ment would be advantageous.544 RERORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. ‘TESTING METHODS FOR STATOSCOPES. ‘PRELIMINARY TESTS. . Before proceeding with the ordinary routine tests, all types of statoscopes are examined fas regards workmanship and for any mechanical defects. A bubble instrunient is inspected for broken and disconnected tubes. A mechanicel type is tapped lightly to determine whether the pointer is loose on the shaft. It is then given a slight rotary motion, so as to deflect the pointer. If serious oscillations occur, the hairspring is too weak or a. part of the lever system is disconnected. ‘The mechenism is then tested for balance by placing the instrument first in ‘ horizontal and then in a vertical position and noting any change in the position of the pointer. ‘Mechanical instruments are tested for leaks in the following manner: The instrument, with its air outlet closed, is placed in a glass bell jar or other container in which it can be observed. ‘The pressuro is reduced until the pointer indicates the maximum deflection, and the instrument, isheld at this pressure for one hour. If during this time the reading of the instrument, decreases, leak is indicated. Care is teken that the temperature of the instrument is kept constant throughout this test. : ‘TEMPERATURE TESTS. ‘Tho thermal insulation of the air chamber of the instrument is tested as follows: Tho instrument is placed in a temperature chamber at —10° C. and kept there for threo hours to Boll jor Borometer O Stetoscope| ‘26 —Disgram of onsatons fr etcg babble trpenateans. ‘Pr. B—DIagrm of coanato for eating mocha! ateope. make sure that it has atteined throughout the temperature of the chamber. It is then quickly placed in e second chamber, at e temperature of +40° C. and read at intervals for two hours. ‘The barometric pressure is also noted so that corrections ean be epplied for change of atmos- pheric pressure. During this test the instrument should not give @ greater indication than would be produced by # change in altitude of 1,000 feet (i. o., a change of pressure of about Linch of mereury), allowance having been made for eny change of atmospheric pressure during the test. ‘The instrument is elso observed for ariy chenge of reading duo to oxpansion or con- traction of the mechanism, with temperature change, and the bubble statoscope for large changes in viscosity of the indicating liquid. sccm pore PRESSURE DIFFERENCE TESTS. These teats aro.conducted in the following manner: In the case of the bubblo type, the instrument is placed in a bell jar which is connected to the berometer and tho vacuum pump as is shown in’ the following illustration (fig. 25). When adjusted and the apparatus is found free from leaks, the pressure in the system is reduced until one end of the bubble is brought to ‘certain mark on the curved gauge. After the pressure has been kept constant for a short time, the reading of the baromoter is noted. ‘The pressure is now reduced until the bubble breaks, and comes back to the same chosen position in the curved gauge. Another reading of the barometer is taken, and the chango of pressure cohputed from the two readings of the berom-ALTITUDE INSTRUMENTS. 545 eter. ‘These observations of the instrament should be repeated about ten times. The average of the readings is computed, and is mown as “the pressure equivalent of the bubble.” ‘This may also be expressed as a change in altitude. ‘Tho mechanical statoscope requires some method by which the outlet to the external air may be closed at-any desired pressure. This was done at the Bureau of Standards by means of a specially designed hose connection as shown in the accompanying diagrams. Figure 26 shows the barometer, vacuum-pump, and bell-jar connections, and figure 27 the de- tailed tube connection. The small rubber tube of the instrument is atteched at (1). Another small rubber tube equipped with a pinchcock is attached. to the copper tubes (2) and (8). After these connections are made and the system is free from leaks, the instru- ment is tested in the following manner: The pinchcock is removed so that the air chamber of the instrument is subjected to the pressure in the testing system, the pump is started, and the pressure reduced to tho desired ‘amount. If there is a resistance to the flow of air in the connections, the statoscope will show a small defection, but the pointer will come to its zero setting when the pressure ceases to change. ‘The pinchcock is now put on the rubber tube at (4), a pressure reading is taken on the barometer, ‘and then the pressure slowly reduced until the pointer of the instrument movas to the first graduation on the dial. At this point the vacuum pump is shut off and as soon as equilibrium is reached the barometer roading is noted. This procedure is continued for each greduation until the end of the scale is reached. ‘The pressure in the bell jer is now increased and observa- tions made at each point on the scale of instrument through the zero point to the end of the descent scale. Check readings are taken by repeating the above procedure. From these obser- vations « table of “pressure equivalents” for each graduation is prepared. ‘TESTING METHODS FOR RATE-OP-CLIMB INDICATORS. ‘TASTING APPARATUS. ‘Tho setup for testing rate-of-dimb indicators is shown diagrammatically in igure 28 and consists of the following equipment: Bell jars and stand—The bell. jars are of glsss, various sizes being used to meet the needs of tho instruments to be tested. ‘The stand is made of a circular steel plate, 20 inches in diameter and one-half inch thick, mounted on threo steel logs. ‘This plate is equipped swith both pressure and electrical connections. ‘Two pressure connections are required for the testing of rate-of-climb indicators, one for an outlet to the pump, the other for the connec- tion to the barometer, since it is necessary to minimize the friction of the ow of air be- ‘ween the bell jer and the barometer so that the barometer will indicate the true pressure change (rate of climb) in the bell jar. The bell jar is sealed to the stand with a soft paste made of paraffin, beeswax, vaseline, and rosin. “Tee mattodafcmnetng ha Droneterio Weel Javon wil abate te pope prema shold be ase. hls naa ould Sagi pte ocr ta now aa hc ana. Pes ween ann We ak 201672335 ‘ro, Diagram of snnations fo testing tot nde,546 REPORT NATIONAL ADVISORY COMMIDTEE FOR AERONAUTICS. Barometer.—Tho barometer used was designed and constructed et the Bureau of Standards and is provided with both pressure end altitude ecalas. The altitude scale is road during tho tests on rate-of-climb indicators; and the rate of climb determined by finding with a stop watch the time required for the mercury to fall a chosen eltitude interval. This altitude interval should not be greater then 1,000 feet since the rate of climb is likely to chango during larger intervals and thus give inaccurate results. By using a large-capecity pump and volumes of about 12 oubie fect, a fairly constant rato of climb can be obtained over a pressuro interval corresponding to several thousand feet of altitude. . Capillary tubes.—The capillary tubes are adjusted s0 es to give the desired rato of climb by grinding to the proper length and by properly constricting the bore. By making up a set of these and using them in various combinations in conjunction with a variable volume, any rate of climb can be produced from 100 fest up to the maximum range of any instrument. Variable volume.—This. consists of a, large glass bottle (B) of about 2 cubic feet capacity connected to the expansion tank (T) and to a second bottle (B’) of tho semo capacity. By varying the level of (B’) water can bo forced to or from the bottle (B) thereby varying the volume of the system which included the calibrating bell jar (J). Sinco the pump (P) is operated at constant speed the same rato the change of pressure (rate of climb) in tho bell jer can in this manner be accurately controlled. When the required volume adjustment hes een made the bottle (B’) is cut off by closing the stopcock (S). Expansion tank.—This is a hot-water boiler (T) (fig. 28) of about 10 cubic feet capacity. which serves the purpose of adding volume to the system. Tt is insulated s0 as not W causo any pressure chatiges in the system due to sudden changes in the room temperature, Vacuum pump.—For ordinary testing a smell pump (P) (Gg. 28) connected direetly to the test chamber is sufficient, but for repid rate of ascent and for more accurate work a large pump and tank are required. ‘Tho ptimp used has sufficient capacity to evacuate 10 cubic foet from 30 to 6 inches of mercury in 10 minutes. [TESTING OF RATE-OP-CLIMD INDICATOR. The tests on rate-of-climb indicators used by the Bureau of Standards aro arranged to dotermine the accuracy of the instrument under the various conditions of fight. ‘Tho main factors which affect the reading of the instrument are lag, changes of temperature, and changos of air density. Preliminary testa—Each instrument is inspected for defeots in workmanship, defective tubes, and leaks. ‘The capillary leak tube should be clean and made of material that will not corrode. Calibration tests. —Tho apparatus is assambled as indicated in figure 28. A chosen rate of climb is produced by starting the vacuum pump end opening the stopcocks connected to the capillary tube which has boon selected to give the desired rate of climb in thosystem. If tho testa require that this rate be definitely specified—tor example, exactly 100 fect por minuto— the levels of the liquid in the variable-volume bottles are adjusted until this rate of climb is obteined. When the spparatus gives the proper rete of climb the system is opened to the air and allowed to come back to atmospheric pressure (care of course being takon that the instru- ment is not damaged by being subjected to a too rapid rato of change of pressure). ‘The pump is now started again, and as the mercury passes a chosen altitude division on the barometer, sey, the 1,000 feet mark, the stop wateh is started. When the mereury passes another chosen mark on the barometer the watch is stopped and the time interval noted. ‘During the above operation another observer simultaneously takes readings of the instrument. ‘From the observed time interval required for the mercury to pass the two chosen altitude divisions the rate of climb is computed. ‘This operation is repeated until the desired number of points on the scalo of the rato-of-climb indicstor sre teated. ‘The results of theso tests are charted by plotting the rate of climb as given by the barometer against the reading of the instrument. See following ourves (figs. 29 to 35) of various instruments tested at this bureau.Colculated’rateafclim® inone hundred feet par minte Calculated rateafclimivinane hundred feet per mnute 12 ' ad Tre cheated calories 1 : a LF | eo a tthrlrlLhLrlL | a 6 Belt, ot g 4 fe ; ; a PrPPyr ty s, lA ALIIEUDE INSTRUMENTS. BAT the ofper fect coloration —. * ~ i g 72 3 eg 7 = Reading otinstrumantin onehundredtee! per mintite Reading of nstrarmentitiane hunted per min. oa, S1—Caletin of German boon yrometer No.1. ‘Fs 2-Catestono¢ Geman baller No.2.548 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS, ___ Tests to determine variation of scale value with altitude —This test is conducted by maintain- ing a constant rate of climb in the test apparatus and making rate-of-climb observations on at [the ofperfechcolibration (fecico 8 5. ® @ iN { Cakulated rateofclimb none hunchedteetper minute (Cae ueaaere earn Te eae 2 24 Reoding ofinstrument inone huntired fiper mintie Fro. 8—Celiaion of Baren of andes ateatlinb neater No. barometer and instrument as in the celibra- tion test, for every 6,000 feet from sea level to 40,000 feet. ‘rom the data obtained in these tests the ratio of the calculated rate of climb from the barometer and stopwatch obser~ ‘vations to the reading from the instrument, are computed and plotted against the altitude Results of tests on various, types sro shown on the following graphs (ge. 36 to 40). Lag teste. ~The theory of the behevior of rato-of-climb indicators shows that the time lag depends both on the altitude and the preestire difference between tho tw ends of the leek tube. condueted to determine the variation of tho lag, both with altitude and deflection. ‘The lag of the instrument is determined in the following manner: A chosen rate of limit is produced in the apperatus. When this has been held constant until the pointer becomes steady (in most cases a minute is sufficient) the instrument is read. and pump ere then suddenly shut off from the system thus 22 2 16 ‘The calibration tank TTT 25 AM 25°C offer Night of kine 25: A eB ehorter Map of sarees TTTTy aa eal oul 2 i | stopping the pressure change in the container. At the same in- stantastop watch isstarted and y then stopped when the reading § of the instrument comes to one- £4 third of its former value. ‘The Jag is the time required in seo- onds. For a constant deflec- >*| tion these readings are repeated ‘every 5,000 feet until 40,000 foet of altitude are reached. ‘The lag is also observed for § a series of rates of climb, thus £ 2 showing varietion with defeo- § tion. 3 Temperature teste A cali- © 8 bretion test is run at a tem- § perature-of —10° 0. to deter- mine the change in slope of tho *| calibration curve with temper- ature. ‘Tho instrument is placed ina temperature chamber where it ia kept for four hours at the Wve temperature to allow the t a ey Reoding ofastrument in one hundred feel per mista 18 20 2532 ‘ho. 34Claton o.oo No.9 ater eh ts, instrument, to come practicelly to the temperature of its surroundings; then » calibration similar to that described above is obtained. The following observations are made during theALTITUDE INSTRUMENTS. 549 progress of the above test. Just efter tho instruments are placed in the cold chamber they ‘are observed to see if they indicate a descent. ‘This shows the efficieney of the thermal insula- tion. After an instrument of a liquid type hes been subjected to e low temperature for two hours the indicating liquid is observed for change in its viscosity, and the mechanical type for change in zero due to contractions in the mechanism. * ue eee tee ead Mitac in Wiocsercts of fel Mra, Varta of wale raat with tage ee 2 4 Fa] : 6 ERLE EL, ; (ae cpo i a ee cane So : i ke : sss . 3 fe i ‘s i sap ea nT a TORT Reading ofinstrumentn one hurrired feel per mite 35 Callen of Geran rateaketintdeatr, mecha tpt a it ° Witide n thousands feet Miler thousond of fet 1, bustin of als val with slttide. Coma tacos Fo. dt-Waraton of wale rise with ite, Geman rate Tanne oe 6 et permis Pies ON Gb ndatr No, ro an pe ste. Inclination tests on mechanical instruments.—All of the above tests’on the mechanical in- truments are made with the diel of the instruments in a. vertical position. A calibration test is repeated with the dial horizontal. If there is a considerable change in the slope of the eali- bration ourve # complete set of tests as described above are made on the instrument with the ial in a horizontal plane.REPORT No. 126. ALTITUDE INSTRUMENTS. PART IV. AEROGRAPHS AND STRUT THERMOMETERS. Br Jom A. C, Wanwea, SUMMARY. ‘This part contains # description of the princijfel types of thermographs and other aoro- graphic instruments, together with a discussion of performance characteristics and methods of testing. ‘Direct and remote indicating strut thermometers are elso included, for whilo these instru- ments are not mechanically self-recording, they are commonly employed for the purposo of proparing curves to show the temperature distribution at successivo altitudes and are tharefore chiefly of interest in connection with acrogrephic instruments. ‘THERMOGRAPHS, In the field of aeronsutios it is often important that a temperature-time record be socured, either in connection with performance tests of aircraft or in obtaining meteorological information of valuo to the flyer. ‘The thermograph is designed for this purpose, and may be used to record cockpit temperatures to which the flight instruments are subjected, or in its meteorological eapacity to rocord the free-air temperature at the ground or at any altitude to which it has boon, lifted by airplane, balloon, or kite. Tho records thus obtained facilitate the application of {instrumental corrections made necessary by the errors of tho instruments due to varying temperature conditions, or provide data for the reduction of altitude determinations as wall as ‘useful meteorological information unobtainsble by other means. ‘All types of thermiographs combine somo form of temperature element and recording device with e timing or clock mechanism in such a manner that the record chert presents a curve of temperature against timo over some definite poriod. ‘The general principles of operation are identical, no matter what tho particular duty of tho instrument may be. ‘TRMPERATURE ELEMENT AND RECORDING DEVICE. ‘Tho commonest types of thermograph depend for temperature indications upon either the Bourdon tube or the bimetallic strip. Of these, the Bourdon tube instrament is perhaps more ‘common and will be considered first. Bourdon tube typé-—Tho ection of the liquid-filled Bourdon tubo as s temperature clement is so well understood that vory little explanation will be undertaken in this paper. Suflce it to say that the affect of varying temperatures upon the curved metal tube of elliptical cross- section, completely filled with a liquid at the proper pressure, and hermetically sealed at its ends, is to cause it to change its curvature and assume varying positions as tho inclosed liquid expands or contracte with temperature changes. Referring to 6g. 1, the Bourdon element will be seam at (A), pivoted at the end of a bracket which holds it outside the instrument case where it may be influenced by free-air conditions. ‘The Bourdon cloment of « typical instrument has an elliptical cross section approximately 560ALTITUDE INSTRUMENTS. 551 33 millimoters slong its major axis by 2 millimotore slong its minor axis, and has a length of 65 millimeters slong its curved edge. Adjusting screw (B) acts upon the tube through the lever farm attached to the pivoted end of the tube and so allows for satisfactory adjustment. ‘The adjusting arm is held in place by the compression spring shown at (B). Tho wire frame below tho tubo serves as a protection from mechsnical injury as the instrument is moved about. At the lower or movable end of the tube and rigidly attached to it is a connecting rod which transmits the motion of the tube through a simple linkage, pivoted near the top of post (©) to'the light spring motal srm (D), which carries the recording pen at its outer extremity. A thumb nut is provided at (G) so that the pressure of the pen on the chart may be properly regulated. The vertical spindle (F) may be moved by the small lever attached at its lower end in the base of the instrument so as to raise the pen from the chart when the temperature record is not desired. Bimetallic strip type —The bimetallic strip principle has long been used in many types of instruments and apparatus. When two strips of motel with different thermal coefficients of expansion are firmly joined by soldering, welding, or otherwise along their entire length, a change "ro 1 Bourdon abeaemogag "Fo. 2 Distale wip termograph in temperature causes a distortion so that the bimetallic strip assumes a shape approximating an aro of ecircle. When the initial temperature is restored, the strip returns to its former shape. Inasmuch as the distortion has a definite and practioally fixed relation (within proper tempera- ture limita) to the existing temperature, the bimotailic strip presents itself as a Suitable element for e temperature-recording instrument. Tt is possible to combine practically any two metels with different thermal cosfficients of expansion to form a temperature element, but in cases where a relatively lange distortion is desired it ia necessary to choose metals whose coefficients differ materially. Bress or bronze with steel or invar make suitable combinations. ‘The ratio of the coefficients of expansion at room temperature is approximately 2 in the case of ordinary brass to steel and 18 in that of brass to invar. These ratios vary over a cozisiderable range, however, according to the alloys employed. ‘The strips are used in the form of helices, U-shaped members, or as straight picces. The advantage gained by the helical or coiled form is that the distortion which takes place under varying temperature conditions produces a direct rotative motion thus making the use of linkages or levers in the recording mechenism unnecessary.552. REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. Such is the case with the American instrument shown by figure 2, in which the helical member is sean at (A). In a certain instrument of this type the strip has the following approxi- Fra 3 Talan cement mate dimensions: length, 355 millimeters; width, 6.5 millimeters; thickness, 1 millimeter. ‘This forms a helix approximately 33 millimoters, in length by 82 millimeters outside diamoter. A spindle coincident with tho axis of the holix cylinder and supported at its oxtremities by Dearing posts (C) holds the elemont in posi- tion and serves to transmit its motion of ro- tation to the light spring metal arm (D) mounted at the end of tho spindle and carrying tho recording pen point at its outer ond. ‘Tho element is provided with two adjust- ing devices indicated by (B) which also con- neot it to its spindle. A still further adjust- ment mey be effected by means of tho sect screw which holds the recording arm in place at the end of tho spindle. ‘Tho pen-point pressure may be varied by the adjusting thumb nut (G) and the pen raised from the record chart by movement of the vertical rod (F) as previously deseribed. Figure shows an interesting Italian adap- tation of the bimetallio strip temperature ele- ment replacing the pressure chambers of an ordinary aneroid type of barograph, thus con- ‘vertingitinto @ thermograph. ‘This particular instrument is ‘provided with two recording pons, the first to give temperaturo records and the second to mark time intervals when acted ‘upon by an electromagnet energized at definite intervals by en intermittent current externally controlled. ‘The record drum is caused to ro- tato by the original barograph clock mechanism, Fie, & Bue te meape A metal tube mounted on the case cover pro- vides for communication with the outer air so that e satisfactory circulation around, the tom- perature element inay be maintained. (Letters reprosent samo elements as in provious doscrip- tion.) . Figure 4 shows a small bimetallic strip (cine-steel) thermograph of German menufec- ture. Several interesting features are incorpo- rated in the mechanism, such as the straight bimetallic clement, the method of balancing tho linkage (see weight at ond of recording arm), and the timing mechanism which is contained in one of the vertical record cylinders. ‘The other cylinder is mounted on » movable base so that the distancebetween the two cylinders may be varied by moving a lever when it is de- sired to remove the chart. ‘The instrument is constructed throughout with a viow to extremo lightness and weighs but 624 grams.ALTITUDE INSTRUMENTS, 558 RECORD CHART AND TIMING DRUM. ‘The record chart and timing drum common to all thermographs ere very similar in design in all ceses. The record chart, in the form of a paper strip suitebly ruled and marked, is held in place upon the rotating timing drum either by a spring strip of metal as shown in figure 1 or by having its ends secured by means of glue. ‘The ordinetes of the chart represent tem- peratures end the sbscissee, time. The Bourdon tube and bimetallic strip instruments are adapted to a comparatively large range of temperatures, and so may be used with charts of varying limits. The inital adjustment is made for each type of chart by means of adjusting serows, as tho indications are compared with those of @ standard. ‘The charts generally provided cover a range of about 50° or 60° of centigrade scale, but the instruments designed for the study of conditions at very high altitudes require a greater range with a comparatively extreme lower limit. A millimeter of the temperature scale ordinarily represents about 1° or 2°. ‘The instruments shown by figures 1 and 2 make a weekly record on a chart about 29 centimeters long, while the German thermograph and French baro-thermo- graph each make 2i-centimeter records in 5 end 6 hours, respectively. ‘The clock mechanism which causes the record drum to rotate at definite rate is usually contsined within the drum itself. It may, however, be mounted in the base of the instrument as shown in figure 2. In this latter disposition the drum may be removed for the renewal of the record chart without disturbing the clock mechanism. By this arrengement the drum is also lightened considerably and friction reduced thereby. When the clock movement is inclosed within the drum, a space otherwise unused is utilized, and the simple handling of the clock mechanism without involving the whole instrament is made possible. Te is important that the driving connections between clock movement and record drum of an instrument for use on aircraft be free from lost motion. ‘This difficulty may be over- come in most cases by careful workmanship in the construction of the parts and in their adjust- ment. In some instances this cause of irreguler drum action has been avoided by making direct connection between the mainspring and drum. The construction of the clock movement es regards the time required for a complete rote- tion is dependent upon the kind of service for which the instrument is to be used. ‘The common meteorological thermograph for recording temperatures at ground level is ordinarily equipped with an eight-day clock movement which carries the drum through one complete turn in @ week's interval, with graduations and markings on the record chart indicating the successive days— while the instrument to be used in the upper atmosphere is designed so us to provide for complete rotation in the maximum number of hours which tho fight may occupy. Certain of the more recent instruments are equipped with clock mechanisms provided with a gearshifting devico which allows the choice of several different rotative speeds. Records of Aights of varying length may thus be made and the timé scale value adjusted so as td suit the conditions. 'THERMOGRAPH COMBINATIONS. . ‘The thermograph is often combined with other instruments to form a single unit record- ing several distinct quantities against time, the records appearing upon a single chart. The barograph or the hygrograph is most frequently found in this combination. Figure 5 shows ‘a baro-thermograph (without case) of French manufacture, the pressure markings being at the lower part of the chart, with temperature record above. The Bourdon tube element is under- neath the base of the instrument where it is protected from mechanicel injury by the cage as shown. (Letters represent same elements as in previous descriptions.) ‘A description of the Marvin, the Fergusson, and of a French meteorograph will be found below. ‘These meteorographs are combination. instruments incorporating the temperature eloment with others.554 REPORT NATIONAL ADVISORY COMMITTEE FOR ARRONAUTIOS, ‘TESTING OF THEEMOGRAPHS. ‘Two principal thermograph tests are required for the determinetion of calibration errors and lag constent of tho instrument. ‘The laboratory installation for this work consists of a properly constructed temperature chamber with glass observation window and oquippod with heating and cooling coils of sufficient capacity to cover the temperature rango of the instrument, ‘under test. A fan system is usuelly arranged to provide proper circulation, which insures ‘uniform temperature conditions throughout the chember. A standard thermometer of known calibration and characteristics is mounted in the chamber, where its readings mey bo comparod ‘with those of the instrument under test. Calibration test—The instrument is first put into propor running order with tho clock mechanism functioning and the recording pen making @ fine lino upon the record chart, The adjusting screws are then regulated to mako the reading correct for the surrounding tempora- ‘ture, ‘The thermograph is next placed in the chamber, which is then closed and tho tost bogun. ‘Tho temporaturo of the chamber is first raised to the upper limit of the instrumont, in two or throo stages of epproximately equal temperature intervals. At each stage the temperature is held constant for a sufficient period to allow the pen to come to rest and for the standard to roach its equilibrium temperature. Readings of the thermograph and standard are taken. ‘Tho thermograph is then shaken and tho reading taken again. ‘The difference in those two readings is the friction error of tho instrument. ‘The heating coils are then disconnected and the temperature allowed to fall. In caso it is desirablo to determine the magnitude of hysteresis effocts, tho descont from the higher temperature may also be made in stagos and the “up” and “down” roadings compared. ‘The same procedure es outlined abovo is fol- lowed for the lower part of tho scale by bring- ing the cooling coils into action, ‘A thermograph calibration eurvo is op EE oe Shown by figure 6. The large departures Trermegraah reoding -"C at the lower end of the tomporature scale aa ‘are duo primarily to the change in the ratio of tho offoctive lengths of the lever arms as the deflection increases. ‘The simplest method of overcoming this error is to determine by oxperiment the deflections corresponding to given temperatures, and to rule the chart accordingly. If, however, the thermograph is adjusted to cover some temperature interval other than that for which the chart is prepared, the ruling will in general show calibration errors, since the adjustment shifts the position of the lever arms with reference to the chart Lag test.—Considerablo lag is found in even the best thermographs. In tho determintion of the lag constant, temperature chamber and standard thermometer are usod as in the above test, The temperature of the air immediately surrounding the thermogreph is first carefully dotermined. The heating coils in the chamber are then connected and the temperature of the chamber brought to e point 10° or 15° warmer than that of tho instrument under test. The thermograph is then quickly transferred to the chamber and careful nota is made of the timo required for the difference between the instrument reading and the chamber temperature to be Additive vorechone ate oy Ao reduced to } times its original value. Tt is in order to simplify mathematical calculations in ‘hich occurs the value ¢ (equal to 2.718), base of the Napetian system of logarithms, that the fraction is customarily taken as 2* As an example, let us assume that the initial difference botwoen taermograph indication and chamber temperature is 10.8° C. The lag constant wouldALTITODE INSTRUMENTS, + 555 4 10.8°C,, that is, 4° of the chamber temperature. It is well to make several observations of this time lag constant with the chamber at various temperatures, exch time bringing the instrument back to its original temperature before making a new test. lock test.—The clock may be compared with a standard timopiece in order to investigate its proper functioning and to determine the errors in rotative speod of the timing drum. This test should bo made with tho record chart in place and with the pen resting thereon as in actual service. Clock mechanism subjected in service to vary Jow temperatures should be given low tem- perature tests. As is the case with other instruments, the mechanism should be warmed after these testa so as to drive out any collected moisture which might otherwise cause corrosion. Tt may be advisable in some cases to conduct additional tests such as thet to determine tho effect of vibration, but this procedure is ordinarily unnecessary. Additional taste —Tn caso tho thermograph is likely to-bo subjected to vibration during use, as, for example, on an airplane, it is advisable to conduct vibration test by mounting the instrument on a vibrating board and determining whether the action of the recording pen is satisfactory. Itshould not show excassive vibration. A tilting test may also be conducted by placing the instrument in various positions and noting the changes in position of the pen. It is sometimes advisable to make pressure test on Bourdon tube thermographs. In properly constructed instrument of this type very little effect is noticed with the decrease in atmospheric pressure which takes place with changing altitudes. In case it is desirable to investigate this point, the instrument may be placed in vacuum chamber in which the pressure is reduced to correspond to the maximum altitude and minimum temperature for which the instrument is to be used. ‘There should be no appreciable change in the indications of a prop- rly filled instrument as the pressure is reduced. if the temperature is kept constant. Care should be taken to see that the Bourdon tube contains sufficient liquid to keep the tbo under pressure throughout the temperature range measured. Otherwise the apparatus may exhibit an erratic and irregular behavior. ‘MBTEOROGRAPES. ‘THE MABYDE KITE METFOROGRAPH, ‘Tho instrument illustrated by figure 7 is the Marvin kite moteorograph, designed by Prof. ©. F. Marvin, Chiof of the United States Weather Bureau. Intended for use in exploring re- gions of high altitude, this instrament is very light in weight (1,188 grams) and makes simul- taneous reoords of atmospheric pressure, temperature, relative humidity, and wind velocity upon suitably divided record sheet mounted upon a timing drum, In this instrument, inclosed in its protective case, is secured in proper position on an aircraft or inside a kite having sufficient lifting power to carry it to the desired altitude. ‘The screening tube seen above the record drum in figare 7 contains the anemometer for air velocity measurement, the temperature clement for temperature measurement, and the hygrograph hair which serves 2s the sensitive element for humidity records. ‘The aneroid pressure element is seen ints position between the sereening tube and record drum. All of these sensitive elamenta connect through suitable devices and linkages with pens resting upon the record chart. Air velocity lement.—Tho air velocity element consists of a small anemometer fon mounted on light bearings inside the forward end of the screening tube. The rotative motion of this anemometer element is transmitted through worm gearing to e cam. A lever connected to the recording pen pivot bears against the cam so that as the Istter comes to a certain period in its motion the lever is pulled down and the pen makes # mark upon the chart. ‘The pen and lever then return to their original positions, and the action is repeated at intervals whose length depends upon the pitch at which the fan is adjusted.556 REPORT NATIONAL ADVISORY COMMIVTEE FOR AERONAUTICS. Temperature element.—This instrument depends upon o bimetallic (bronzo-invar) strip for the temperature records. ‘The element, 25 millimeters in width and bent into circular form, is mounted in the soreening tube, end its free end, which changes position in a definite and regular ‘manner with temperature variations, is connected to the recording pen arm by means of a simple Tinkage. An adjustment is provided for changing the initiel position of the pen upon the rocord chart, and another device makes it possible to vary the sosle value as desired. ‘Humidity clement.—Similar to inost devices for recording indications of relative humidities, this meteorograph employs a series of human bairs which have the property of changing their length with varying conditions of moisture, Instead of combining the hairs in tho usual bundle form, the designer has mounted them separately, which enables the hairs to come into equi- librium with the surrounding air more promptly. " ‘Two sets of hairs are mounted longitudinally in the screening tube upon suiteble fixtures provided with the necessary adjustments, and the clement is connected by direct linkage to tho recording pen. Inasmuch es the change in Jength of the hairs is not Tinear with respect to the humidity change, it is necessary to provide « special scale for tho individual clement em- ployed. Adjustments are provided for varying tho scale value and for controlling the initial position of tho record- ing pen upon the chart. Pressure element.—Two nickel-plated steel aneroid chambers provided with internal stocl springs aro em- ployed as the pressure sensitive clement of this instru- ment. In the photograph they are clearly seen, mounted. betwoon the screening tube and record drum. A suitable linkage connects the aneroids with tho recording pen 80 that the latter traces the pressuro curve upon the chert with expansion or contraction of the flexible steel ancroid chambers as the external ai pressuro changes. A small quantity of air ellowed to remain within the aneroid chambers is intended to compensete in part for deffcc- tion of the element due to tomperaturo variation. A. Dimetellic strip in the connecting linkage is also provided for this purpose. ‘The scale value may be varied by ad- justment of the linkege. Record drum and chart—The record chart is properly ruled to receive the traces of the pens connocted with the four elements, ‘The wind curve comes at the top of tho : sheet, with those for temperature, pressure, and humidity occupying successive spaces over the ruled chart.. ‘The latter is held upon a removable timing dram with clock movement inclosed. The drum makes © complete rotation in an eight-hour eriod. . Poa Tasirument housing—Th protective housing (shown in place in the illustration) slips over the instrument eo that no partis exposed excepting the screening tube which contains the several sensitive elements. ‘Two bakelite strips insulata the tube from the case. A mice observation window is provided for the inspection of the recording elements when the housing is in place. “THE FERGUSSON METEOROGRAPH. One of the most interesting of recent developments is the meteorograph by S. P. Fergusson, meteorologist of the United States Weather Bureau. It is essentially » modified baro-thermo- hygrograph designed with a view to extreme lightness, thus making it possible to carry it to great altitudes by means of sounding balloons. Tho instrument with its caso (soe fig. 8) weighs but 180 grams, as compared with 400 grams, the weight of the next heavier instrument of its type. In addition to its lightness, it possesses several other important advantages%as noted in the briet description which follows:ALTITUDE INSTRUMENTS. 557 Tomperature element.—The brovze-inver bimetallic strip temperature element upon which this meteorograph depends for temperature records is mounted in the vertical screening tube at the right side of the instrument where it is exposed to the circulating air. The mounting is 80 designed and arranged as to avoid the effects of radiation end conduction from surrounding parts. The strip is bent into circular form and its free end is connected by a simple linkage to the recording stylus which rests upon the record drum. One millimeter of the temperature scele represents a change of 2° 0. Humidity element—The bumidity element is composed of six or eight series of human hairs of three strands each. They are mounted upon suitable fixtures and extend vertically inside the sereening tube. Tension on the hairs is maintained by a flat spring, one end of which is connected to the recording stylus. Pressure element.—Perhaps the most interesting feature of this instrument is the method {for obtaining a comparatively open scale for the pressure records which may cover practically the entire range of atmospheric pressures. Inasmuch as the pressure scale becomes greatly contracted at extremely high altitudest thus making the records uncertain, it is desirable that ‘special arrangement be made in the instrument design s0 4s to compensate for thie characteristic. ‘The pressure clement itself is a common form of ex- hausted Bourdon tube with ono end securely fixed to the i strument base and with its movable extremity acting upon the recording stylus through a linkage of ingenious design. Tt is this linkage which cares for the open-scale feature mentioned in*the preceding paragraph by causing the recording stylus to make two traverses of the record chart. ‘The first or up- ward traverse covers a certain range in altitudes and the sec- ond or downward trace continues the record to the upper limit. It is elso possible by the adjustment of this linkage to vary the scale so that « certain upward motion of the stylus has a value equal to twice that of the downward motion in order thet small pressure chenges above 10,000 meters may bbe determined with greater precision then would be possible with a uniform scale. Record drum and chart.—Tho record drum seen at the left of the illustration surrounds the clock movement which is mounted upon the base plate. ‘The clock causes the drum to tum through one complete rotation in one hour end is so designed. and connected to the latter that the clock may be rewound by turning the drum backward; the number of turns depends upon the mmber of rotations desired for the record. ‘The chart or record sheet is made of very thin sheet aluminum (0.03 millimeter in thickness) wrapped once around the drum and with the ends secured by a special lock joint. In making the instrument ready for service a leyer of smoke is deposited upon the aluminum chart by holding it Gm place on drum) over @ camphor or kerosene flame. ‘Tho recording styli rest upon thissmoke‘ilm and make fine lines es the drum rotates. In addition to the is styli mentioned above, there is « fourth one which is fixed so as to mark 4 base line at the bottom of the chart. A stylus lifter is provided to raise the markers from the chart when no record is desired. Finally, the record may be fixed or made fast, after it has boon obtained, by applying a suitable preperation’ (whito chellac and glycerin) which hardens the film. ‘The record is examined and evaluated by tho uso of a transparent scale suitably divided. ‘The designer of this instrument has endeavored to avoid complication so that the parts may be easily and economically manufactured and assembled. The entire instrument, in- cluding the case, is constructed with @ view to compactness and lightness without loss of rigidity. The outside length of the instrument is 210 millimeters, the height 90 millimeters,558 ‘REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS. and greatest width 85 millimeters. The clock drum is 80 millimoters in height, 57 millimeters in diameter, with a time scale of 8 millimeters per minute. ‘MRENcH DARO-THERMO-YGROGRAPH, ‘Figure 9 shows a view of a baro-thermo-hygrograph of French design. The instrument is similar to the baro-thermograph shown in figure 5 of this paper, with the exception that in this instrument the hygrograph element is added. Temperature element.—The Bourdon ‘tubo temperature clement is mounted underneath the instrument base, where it is protected from mechanical injury by a shoet-metal framework, A simple linkage connects it to the recording pen arm resting upon the uppermost section of the chart, which is ruled to cover a temperature range from minus 30° C. to plus 30° 0. in a 00 millimeter vertical direction. "Fo. 0—Preachbaotensohyereap The usual form of hair hygrograph element is employed. A bundle of approximately 25 hairs is mounted upon fixtures upon the right-hand vertical framo soction of the instrument. Tension adjustments are provided. A small hook looped over tho hairs midway between the supports is connected with a pivoted lever which carries an arm shaped in the form of e cam and resting upon a second pivoted cam-shaped arm which transmite ite motion to the recording pen. These cams aro held in contact by a light spring in tonsion, This arrangement converts the changes of hair length so as to make the rocording pen movo- ment linear with respect to humidity changes. This makes possible the uso of an evenly divided humidity scale, which is seen at the center of the chart. This section is 60 millimeters in height and is divided into 60 equel spaces. _ Pressure element.—Aa in the baro-thermograph, @ double aneroid clement is usod. ‘This is connected through a simple linkage with the recording pen, which moves over the lowost portion of the chart and covers a range from 760 millimeters to 150 millimaters in a height of scele of 60 millimeters. - Record drum and chart.—The record drum, which contains the clock mechanism and upon which the chart is held in place by a spring clamp, has a height of 208 millimeters and a diamoter of approximately 67 millimeters. It makes a complete turn in a period of eight hours. The chart is 190 millimeters in heigh Instrument dimensions —This baro-thermo-hygrograph complete with case weighs about 1,456 grams. It stands about 30 centimoters high, has a width of 10 centimeters, and a length of approximately 22 centimeters.AUTITUDE DYSTRUMENTS. 559 STRUT THERMOMETERS, As previously outlined in this paper, the proper reduction of aircraft performance date requires the uso of figures showing fres-air temperatures obsarved at various altitudes during flight. ‘The strut thermometer, so called because of its usual location on the strut of en air- plano whero it may be affected by free-air conditions, is most often used for this purpose and for similar duty in other experimental flight. ‘The two principal types will be considered in this paper. RESOTE-INDICATING LIQUID-EKPANSION TPE. ‘The liquid expansion principle which forms the esis of operation of one of the common types of airplane engine thermometer is also used in strut-thermometer construction. In. fact the same arrangement and mechanism with slight modifications mey be used interchange- ably in either capacity. ‘The liquid-flled bulb with its radiation fins (Sg. 10) is mounted in a Yoo 18 Stra nema. suitable position on the aireraftfusually on a strut or on the Ianding gear. An increase in the temperature of the bulb causes the confined liquid to expand, which in turn produces a motion of the Bourdon element in the indicator with which the bulb is connected by means of an armored capillary tube of fine bore. ‘The Bourdon tubs is formed in several concentric helical coils, and its movable end connected to the indicating pointer through a bimetallio-trip heli- cal coil. ‘This bimetallic strip member is provided. to compensato for the changes in the tem- perature of the gege and tubing. The strut bulb should be shielded from the direct rays of the sun by some arrangement which will permit satisfactory air circulation around the bulb. A complete discussion of this type of instrument may be found under the titlo “Thermometers for Aiferaft Engines,” Part III of Report No. 129. In order to allow for # more open soale, the dials of strut thermomoters have a greater diameter than those of the engino instruments, and the scale covers a rango of about 80° C., with —40° or —50° as lower limit. An instrument of this type recently examined hed a dial560 JBPORT NATIONAL, ADVISORY COMMUTERS FOR AERONAUTICS. 10 contimetars in diameter, with a sufficiontly open scale to allow for readings to within one or ‘vo tenths of a degree. Suitably designed thermometers of the vapor-pressure type inay also be readily adapted for uso as strut thermometers. {MQUID-COLIMIN GEASS THRE. ‘The second type of strut thermometer shown in the photograph is a liquid-column glass thermometer of ususl form, designed for the proper temperature range end so mounted in its backing as to make it readily adaptable to use on an airplane. This instrument is ordinarily fastenod to a strut of the airplane where it is subjected to an unobstructed circulation of freo air and is onsily observed by tho test pilot. ‘The bulb is surrounded by a brass case intended to protect it from mechanical injury and also to shicld it from tho direct rays of the sui ‘The wooden backing of the giass thermometer tubo is so shaped as to accomodate itself to the strut form, and felt strips are provided to avoid marring the strut when the instrument is strapped in place. ‘The soale is graduated and proportioned 50 as to make possible the correct reading of the instrument to within about 1° ©. from the pilot’s position. A detailed descriptive treatment of thermometer testing methods and apparatus may be found in Part III of Report No. 129, under the title “Thermometers for Aireraft Engines.”