Under ordinary conditions a pilot is able by his senses of sight, hearing, and feeling to know just about how his plane is performing without focusing his attention on the instrument panel. In fact, for an experienced pilot, most of the instruments serve chiefly to check his estimate of the ship’s performance. The more experienced the pilot, the more independent he can be of his instruments, generally speaking, but no amount of experience will enable a pilot to determine how far he is above the surface of the earth. It may be considered one of the hazards of aviation that the one instrument, the altimeter, for which no amount of experience can substitute, should be subject to so many corrections. An understanding of the altimeter and a few thumb rules pertaining to its corrections should enable a pilot to have sufficient altitude at all times to clear in safety the high ridges and tall trees which frequently give trouble on cross-country flights during thick weather.
The aircraft altimeter used for determining altitude is essentially an aneroid (non-liquid) barometer refined and more ruggedly constructed and with the dial graduated in units of altitude instead of pressure. The average relation between altitude and barometric pressure has been determined by the Bureau of Standards whose figures are used in calibrating all altimeters. The zero of this standard corresponds to a normal barometric pressure of 29.92 in. of mercury and a temperature of 50oF., but the dial of the altimeter is so constructed that any elevation or barometric pressure may be taken as zero at the discretion of the pilot.
The altimeter measures height above a landing surface in a roundabout manner since it cannot measure linear distance directly. What it actually does is to show the difference between the force exerted by the height of the air column at the point of take-off and the same force above the strata in which the plane is flying. Because two air columns of different heights exert different pressures, the forces acting on the altimeter would be different. Atmospheric pressure (as well as any other) is equal to force divided by area.
P=F/A (Pressure=Force/Area)
By Newton’s second law of motion, force is equal to mass times acceleration. Since F=ma, or mg, we may write:
P=mg/A
Now density (d)=mass/volume, or d=m/v, or m=dv
Substituting for m in (2) we may write:
P=dvg/A
But since the volume is the product of three dimensions we may write: Area x height = V
Again substituting and using a unit cross-sectional area we have:
P=dhg
If, for the moment, we eliminate the changes in density of the atmosphere and the difference between the gravitational acceleration at the surface and at points above the earth, we may write:
P=h
Therefore, if an instrument, such as the altimeter, measures differences in pressure, it also measures differences in height or altitude. It also follows that at high altitudes the pressure is less since the air column above is shorter than at lower altitudes and has less weight. It has been found that this pressure reduction amounts to about 0.1 in. of mercury for every 90 ft. of change in altitude. Furthermore, because of its principle of operation, the altimeter is independent of the force of gravity and latitude.
Based upon the above fundamentals, the mechanics of the altimeter are simple to understand. Referring to Fig. 1, E, the primary element of this instrument, is a metal diaphragm or cell which has been evacuated and sealed. Its sides are corrugated to give the walls increased rigidity. In altimeters used in the Navy the air pressure is balanced by the tension of the sides of this vacuum chamber itself, while in commercial types an external spring is used, the tension of which varies with atmospheric pressure. Changes in pressure thus give rise to elastic reactions of the cell. These are transmitted by means of a linking-up mechanism to the pointer on the face which indicates altitude. The long lever, L, of the linking-up mechanism is a bimetallic construction which compensates for changes in instrument temperature. A watch which is not compensated for heat and cold will gain in winter because the stiffness of the main spring is increased with decrease in temperature and, conversely, it will lose in summer because the main spring exerts less force because of expansion due to increase in temperature. Similarly, the walls of the capsule lose some of their rigidity with an increase in temperature or they become less elastic as the temperature decreases, but by employing the bimetallic strip L the linking-up ratio is not changed. Most altimeters are compensated in this manner for temperature changes which are most important at zero altitude.
If a plane is used for purely local flying, the readings of the altimeter may be accepted as approximately correct. If cross-country flights over diversified terrain are to be made, certain corrections of the altimeter readings are necessary in order that they may be dependable when effecting a landing. In general, the readings of pressure or barometric altimeters are subject to three important corrections: (1) for the difference in elevation between the fields of departure and destination, (2) for varying barometric conditions during the flight, and (3) for error due to changes in the mean temperature of the air column.
An altimeter set at zero at the takeoff will indicate during the flight altitudes above the field of departure, for by setting the dial to zero the pilot has established a barometric reference plane above which all heights are measured whether plane above which all heights are measured whether the plane flies over areas higher or lower than the altitude of the field of departure. In some cases significantly large error may be involved when the altimeter is set to read zero for the pressure existing locally. Suppose, for instance (see Fig. 2), a plane takes off from the Naval Air Station, Pensacola (elevation 4 ft.) on a cross-country flight to Atlanta (elevation about 1,004 ft.). If the dial of the altimeter is set on zero at the time of the take-off, it will indicate that the plane is 1,000 ft. in the air when it lands at Atlanta, provided other factors remain constant. Such error cannot be disregarded, especially if the weather is thick or a night landing necessary. Hence many pilots find it safer to set the altimeter before taking off the correspond to the elevation or barometric pressure of the point of destination rather than that of the point of departure. In such a case greater dependence may be placed upon the altimeter when making a landing. If a blind landing becomes necessary en route, the pilot must then make a mental calculation of the difference in elevation of his reference plane (point of destination) and the area in which he intends to land. Mail pilots who regularly cross high ridge of mountains often set their altimeters to read zero for the highest peak in the range, knowing that any reading above zero will give them ample altitude to clear all obstructions.
A thumb rule to cover errors due to differences in elevation between the fields of departure and destination may be stated as follows:
When flying from lower to higher elevations, the altimeter unless corrected will overread by an amount equal to the difference in elevation between the two fields.
This rule is illustrated in Fig. 2 and could be stated conversely but, using the example cited, the plane would be actually 1,000 ft. in the air when the altimeter read zero with no disastrous results. The altimeter error is equal to the pressure exerted by the height of the air column between A and B. Bearing in mind that increase in pressure gives decreased altimeter readings, the addition of the pressure column AB would bring the reading to zero at Atlanta. The error is that due to subtraction of this column AB from the total atmospheric column produced by the change in elevation.
The second type of correction is necessary because of varying barometric conditions during the time of flight. It is a generally known fact that the barometric pressure does not remain the same at any given locality. In addition to the diurnal variation of the barometer, there are migratory high- and low-pressure areas which continually trek across most parts of the earth. The United States is situated in the region of the prevailing westerlies where these high- and low-pressure areas follow one another in rapid succession. In consequence, there are often marked and rapidly changing pressure conditions within short distances. When a pilot sets his altimeter on zero, he establishes not only an elevation barometric reference plane, but also a pressure datum plane. Any changes in barometric conditions will cause the altimeter to read incorrectly, for once the reference plane is established, all altitudes are measured thereafter from it since the altimeter cannot adjust its zero reading to the changing barometric conditions. Suppose a plane took off from New Orleans for Pensacola on the afternoon of March 5, 1932. The pressure at New Orleans was 29.62 in.; at Pensacola it was 29.38 in. If the altimeter were set to zero at New Orleans, the plane would have been 216 ft. in the air by altimeter when it landed at Pensacola, certainly an amount which cannot be disregarded if the weather is thick or a night landing anticipated. On the afternoon of March 21, 1932, sea level pressures along the Gulf Coast were as follows: New Orleans, 29.56 in.; Pensacola, 29.66 in.; Tampa, 29.86 in. Under such a pressure distribution the altimeter of a plane flying from Tampa to New Orleans (see Fig. 3) would have had an error of 270 ft. when it reached the latter. It is true that these pressure conditions occurred during bad flying weather. This fact, however, serves to emphasize the importance of the error, since poorest visibilities and lowest ceilings usually accompany rapid pressure changes, giving maximum altimeter errors at the very time when accurate altitude readings are most desired.
Some altimeters are fitted with an auxiliary scale so that the pointer may be set for the barometric pressure at the field of destination. This altimeter is satisfactory provided there is no necessity for intermediate landings. In its use, care must be exercised not to set the dial for the pressure reading indicated on the weather map because these weather reports have been corrected for sea level conditions. Setting a synoptic weather map pressure on the altimeter would correct for differences in pressure due to barometric conditions, but there would be no compensation for elevation, and the latter is usually the most important, although the former must not be neglected.
During long cross-country flights the pressure may be changing, often as much as 0.02 or 0.03 in. per hour. Thus during a 5-hour flight in summer, when the barometer is varying at such a rate, an additional error of 90-125 ft. would be introduced. In winter, when cyclonic disturbances are more active and better developed, the uncertainty often amounts to as much as 300 ft., and under unusual conditions may be three times as great.
The thumb rule for altimeter errors due to differences in barometric pressure between the points of departure and destination introduced by meteorological conditions may be stated as follows:
Caution is advised when flying to a destination where the pressure is lower than at the point of departure, for then the altimeter will indicate that the plane is in the air when it has made contact with the ground.
Just before arrival at an airport equipped with radio and meteorological service, the pilot may request the actual atmospheric pressure at the altitude of his field of destination. By using this pressure for a zero setting of his altimeter, he establishes a new barometric reference plane which will make three necessary corrections: (1) the difference in elevation between the fields of departure and destination, (2) the differences in pressure due to the meteorological distribution, and (3) the varying barometric conditions during flight. For example, suppose a plane en route from Pensacola to Atlanta and that the pressure at Pensacola at time of departure was 30.00 in. Just before landing at Candler Field, Atlanta, the pilot, being informed that the local pressure is 29.02 in., sets 882 ft. on his altimeter as a new zero reading, thus making the aforementioned corrections.
In addition to corrections for pressure variations, a correction for variation from the standard air temperature of 50oF. must be applied in order that the altitude may be determined more accurately. The atmosphere is not of uniform density. Neither is the density constant. The addition of water vapor makes it less dense, while a decrease in temperature increases its density. The pressure acting on the altimeter element at any given time is that due to the force exerted by the column of air above it. IF the temperature of this air column is increased, it will expand and the altimeter must be taken to a higher altitude before it will read the same as at the original temperature (remember that reduction in pressure gives positive altimeter readings).
The error due to changes in the mean temperature of the air column is illustrated in Fig. 4. Let A50B represent the height of an air column at standard temperature of 50oF. If the mean temperature is increased to 80oF., this column expands and becomes CA80B, the increase being equal to CA80. An altimeter in a plane at level A will not read 10,000 ft., because it must still move
through height A80C, the amount of the expansion above the 10,000-foot level, but will read 612 ft. less than 10,000 ft. according to Table 52 of the Smithsonian Meteorological Tables. The correction would therefore be plus 612 ft. If the mean temperature of the air were less than 50°F., say 40°F., the air would then be more dense, and the force necessary to cause the altimeter to read 10,000 ft. would be obtained when the plane reached N. But the established height by the standard atmospheric column is at A40. When the altimeter reaches A40, it will indicate an altitude too high by an amount equal to A40N. Therefore, for mean temperatures below 50°F., the correction must be subtracted, the Smithsonian Meteorological Tables giving 204 ft. as the correction for 10,000-foot altitude for a mean temperature of 40°F. Thus the altimeter would read 10,204 ft. at Ai0 which would be 204 ft. too high.
Thumb rules for errors due to increase or decrease in the density of the air column produced by changes in temperature may be stated as follows:
(1) To obtain a desired altitude, the pilot must fly higher by altimeter in winter; lower in summer.
(2) An altimeter will underread in summer; overread in winter.
(3) In summer the temperature correction will be positive; in winter, negative.
The importance and magnitude of this error may be shown best by examples. On March 11, 1932, the ground temperature at Pensacola was 41°F., and the mean temperature of the air column computed from an aerograph flight was 30°F. The computed altitude for the flight was 11,600 ft., the altimeter in the front cockpit reading 12,200 ft. and that in the rear 12,150 ft. Using the mean of the readings gives an error of 575 ft., almost all of which was due to deviation from the standard air column. At the same hour the surface temperature at Atlanta was 28°F. Assuming the same temperature lapse rate at Atlanta as occurred at Pensacola, 2.1°F., per thousand ft. (decrease of temperature with altitude obtained from aerograph flight), the altimeter of a plane flying between these cities would have an error of 35 feet. This is but a small correction compared to those introduced by changes in elevation and fluctuating pressure conditions, but the pilot must realize that the deviation of the mean temperature of his air column from the value used in altimeter calibration is still another source of error that cannot be entirely disregarded.
A thumb rule for the amount of this temperature correction depends upon the deviation of the mean temperature of the air column from 50°F., the magnitude of which amounts to 10 ft. in every 1,000 ft. for every 5°F. by which the temperature varies from the normal. Applying this rule to the aerograph flight just mentioned, the correction would be 50°F.-30°F. = 20°/5 = 4x10 = 40ft. per 1,000 ft. If the altimeter reading were 12,175 ft., the error would be 40 x 12,175/1,000 = 487 ft. Since the mean temperature is below 50°F., this is subtracted from the altimeter reading, but if the temperature were above 50°F., it is added.
The Kollsman instrument, designed by the Bureau of Aeronautics and used in the Navy, is a single-capsule design, directly connected to the linking-up mechanism. In types previously used, the walls of the air-tight evacuated metallic capsule were prevented from collapsing by a large curved external spring, one side of which was secured to the base of the instrument and the other to the face of the capsule. If the pressure were decreased, the spring expanded the walls of the diaphragm, the movement being transmitted through a knife-edge to the linking-up mechanism. The lag in this type was often large, not only because changes in temperature produced a varying spring tension. By eliminating the spring and the knife-edge bearing, a simpler mechanism was evolved, the lag of which was less than in the exterior spring type with more moving parts.
To check the amount of lag under service conditions, tests were recently made at the Naval Air Station, Pensacola, with two F6C-3’s (fighters) equipped with Kollsman altimeters. One plane flew constantly at an altitude of 2,000 ft., while the other dove several hundred ft., zoomed above the first, maneuvered violently, and returned to the 2,000-foot level. The altimeter in both planes gave identical readings. Next, the maneuvering plane climbed to 4,500 ft., dove 160 knots, and pulled out on a level with the other plane. Again both altimeters read an even 2,000 ft. Then, the maneuvering plane dove at maximum speed from 5,000 to about 1,000 ft. On passing the reference plane at 2,000 ft., the altimeter of the diving plane read 2,000 ft. After a quick zoom back to the 2,000-foot level, a similar reading was indicated. Such tests would tend to show that lag in altimeters of this type is so small that it is negligible for all practical purposes.
There are additional altimeter errors. The gradual release of the internal stresses set up in the pressure element and in the bimetallic strip during manufacture gives rise to secular errors, often called the zero shift error. Over a period of time, the change in the elastic system of the instrument introduces a deflection error known as drift or creep. Friction, if not excessive, is usually offset by vibration of the plane, but because vibration varies with different types of planes, the error is hard to predict. It must be realized that the stress set up in the delicate parts of an altimeter from vibration or by a hard landing are large, and that if some suspension device or vibration-absorbing material were used, the accuracy and life of the instrument would certainly be increased.
In general, altimeter errors may be classified in two groups: (1) instrumental errors: (2) errors produced by the barometric method of measuring altitude. If the performance of an altimeter were limited to the first class of errors, doubtless they could be eliminated by further developments in design, so that improved altimeters could be relied upon absolutely when making a blind landing. Inasmuch as it is subject to large barometric corrections which no automatic mechanical device can eliminate, the pressure-type altimeter with its inherent defects must eventually be discarded for one which will indicate at all times the altitude of the aircraft with respect to the terrain over which it is flying. Whether an instrument based on electrical or on sonic principles will supersede the present altimeter depends upon the research activity of the various scientific groups working on the problem. However, as long as the aneroid altimeter remains as the most generally used instrument, pilots must at all times remember the two most important thumb rules for altimeter corrections:
(1) I may be high or
I may be low,
My altimeter
Does not know.
(2) I may be high,
I may be low,
I must correct
Before I know!